Professional Documents
Culture Documents
DISSERTATION
of the University of St.Gallen,
School of Management,
Economics, Law, Social Sciences
and International Affairs,
to obtain the title of
Doctor of Philosophy in Management
submitted by
Tobias Schlager
from
and
The President:
Of course, I want to thank my supervisor, Prof. Dr. Peter Maas, for his support
and guidance and for providing extraordinary working conditions at the Institute
of Insurance Economics. Peter always encouraged me for my work and assisted
whenever possible. I am also very grateful to my co-supervisor, Prof. Dr. Andreas
Herrmann, for his interest in my dissertation and my skills and for providing me
with excellent feedaback and creative ideas. After all, writing this thesis was a lot
of fun and all my friends, especially those from Innsbruck, have contributed to
this.
Tobias Schlager
Vorwort
Obwohl ich hoffe, dass jeder, der mir beim Verfassen dieser Dissertation
geholfen hat, wei, wie wichtig seine Untersttzung fr mich war, mchte ich
die Menschen noch einmal ausdrcklich erwhnen, die mich besonders motiviert
haben. Zu allererst wrde ich gerne meiner Familie danken. Jedes Familienmit-
glied hat mich auf seine Weise und zu jeder Zeit inspiriert, von meinem ersten
Schultag an bis zu meinem letzten Tag als Student. Immer an mich glaubend
und mir verdeutlichend, dass man nie aufgeben darf, war der grosse Mosaikstein,
den meine Mutter gelegt hat. Gleichermassen hat mein Vater dazu beigetragen,
indem er mich Naturverbundenheit gelehrt hat und das Verstndnis, Dinge auf
das Wesentliche reduzieren, um einen klaren Kopf zu bewahren. Ich bin immer
noch begeistert von seinen Skiknsten und hoffe, bald mit ihm fahren zu knnen.
Mein Bruder, ein stets analysierender Mathemagician und einer meiner besten
Freunde, hat nicht nur direkt Abschnitte dieser Dissertation mit mir besprochen,
er hatte auch immer ein offenes Ohr fr alle meine Probleme. Meine Grosseltern,
die diese Zeilen leider nicht mehr lesen knnen, sind an dieser Stelle nicht zu
vergessen. Ich verdanke auch ihnen sehr viel und denke oft an sie. Von ganzem
Herzen ein aufrichtiges Dankeschn an meine gesamte Familie.
Natrlich mchte ich auch meinem Betreuer, Prof. Dr. Peter Maas, fr die
hervorragende Untersttzung und fr die tollen Arbeitsbedingungen am Institut
fr Versicherungswirtschaft danken. Meinem Zweitbetreuer, Prof. Dr. Andreas
Herrmann, bin ich fr das Interesse an meiner Dissertation, das Vertrauen in meine
Fhigkeiten und fr das exzellente Feedback sehr dankbar. Abschliessend mchte
ich noch erwhnen, dass mir das Schreiben der Dissertation eine Menge Spass
gemacht hat und alle meine Freunde - insbesondere jene aus Innsbruck - dazu
beigetragen haben.
Tobias Schlager
iv Contents
Contents
Contents iv
List of Tables ix
Summary / Zusammenfassung x
I Introduction 1
1 Research perspective 1
2 Theoretical framing 2
3 Practical relevance 4
4 Objectives 5
BIBLIOGRAPHY 11
6 Discussion 32
8 Future research 36
BIBLIOGRAPHY 40
2 Conceptual development 54
3 Hypotheses development 58
3.1 Economic value . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.2 Development value . . . . . . . . . . . . . . . . . . . . . . . . 60
3.3 Social value . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.4 Diversity value . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.5 Reputation value . . . . . . . . . . . . . . . . . . . . . . . . . 63
4 Empirical study 64
5 Research findings 66
5.1 Current employees . . . . . . . . . . . . . . . . . . . . . . . . 68
5.2 Potential employees . . . . . . . . . . . . . . . . . . . . . . . . 71
6 Limitations/future research/discussion 72
8 Conclusion 76
BIBLIOGRAPHY 77
BIBLIOGRAPHY 129
1 Introduction 141
5 Discussion 168
BIBLIOGRAPHY 177
List of Figures
Introduction 1
1 The Dissertations Focus on the Customer. . . . . . . . . . . . . 4
2 Localization of this Dissertation. . . . . . . . . . . . . . . . . . 8
List of Tables
Introduction 1
Summary
Zusammenfassung
Um dies zu erreichen, ist die Dissertation wie folgt strukturiert. Teil I dient
der Einfhrung in die Themenbereiche der Dissertation und verdeutlicht deren
konkrete Ziele. Teil II vertieft die theoretische Perspektive anhand einer konzep-
tionellen Aufarbeitung der Bedeutung von dominanten Logiken fr Customer
Value. In Teil III wird dies aufgegriffen und ein konzeptionelles Framework fr
Service Branding entwickelt, welches empirisch getestet wird. In den beiden
letzten Teile, Teil IV und V, werden die theoretischen Grundgedanken der Logiken
implizit verfolgt, indem ein Forschungsbereich vertieft wird, der hauptschlich auf
Interaktionen basiert: die Multichannel- oder auch Interaktionswege-Forschung.
asdf
1
I Introduction
The following introduction outlines the research perspective that the author of
this dissertation takes in and picks up current discussions in marketing literature
related to interactions. Building on this, the theoretical underpinning, the practical
relevance, and the objectives of the presented dissertation are briefly delineated.
The main body of the dissertation unfolds in the Parts II to V, which pertain
to individual articles that commonly emphasize the dynamic interactions that
customers engage in.
1 Research perspective
2 Theoretical framing
companies: Albeit customers and companies have balanced roles when interacting,
the central focus is placed on the customers point of view.
The above-mentioned focus on interactions forms the common basis for the
research questions of the four articles presented in this dissertation. The central
position of the customer is graphically illustrated in Figure 1. Around the customer,
an exemplary set of actors with whom the customer may interact is sketched,
illustrating the various interactions studied in this dissertation.
4 I I NTRODUCTION
3 Practical relevance
In fact, interactions have not solely been a key topic in the marketing literature in
the past decade but also their practical relevance is paramount. How to interact
with customers constitutes a central pillar of most companies strategy. Whereas
companies such as Dell explicitly consider a specific type of interactions (online
interactions) in their strategy, the question how the customers may best connect
with the company is at least among the major topics of the agenda of other
companies.
with multiple sources just for gathering information before making a purchase.
These figures are backed by a cross-industry study of Google (2009) finding that
only 40 to 60 percent of customers (depending on the industry) consult the same
interaction point for searching and purchasing. What is more, radically different
ways of interactions that consumers can engage in are created by new media. For
instance, whereas traditionally word-of-mouth (WOM) represented an important
way to gather information about a product, service or company, it is nowadays the
e-word-of-mouth (eWOM). Coevally, the range of customers reached by positive
or negative eWOM significantly exceeds the ones of traditional WOM, far beyond
what was imaginable a decade ago.
It is needless to say that the trend towards using the multitude of possibilities
for interacting with companies is still continuing. Along with the technological
developments and the simultaneously changing customer behavior it is imperative
for marketing practitioners to rethink old-fashioned strategies and practices.
4 Objectives
Accordingly, Part II has the objective to shed light on the influence of dominant
logics on the customer value concept, which has so far largely remained under-
investigated. Revisiting the creation of value, the part includes management
literature. Related to this, the objective of the Parts III-V is to contribute to
6 I I NTRODUCTION
specific topics from a dominant logics perspective. Particularly, Part III reflects
the importance of the employer brand for service branding integrating human
resources and marketing perspectives. To date the literature has neglected the
importance of favorable employee perceptions and the creation of a service brand.
The objective is to develop a profound understanding of the network of interactions,
where especially the customers interactions with various actors (i.e., the company
and the customer contact employee) are decisive. The last two parts deal with
the consumers behavior when interacting with companies, or more precisely,
consumer multichannel behavior. The first objective of both is to enhance literature
regarding the consumers interactive behavior with companies in the sense of what
Neslin et al. (2006) understand when referring to channels: Ways of interaction
instead of solely one-way communication. Thus, both articles emphasize the
consumers proactive role in interactions. While Part IV follows this notion by
establishing an international consumer segmentation, which includes emerging
markets, that founds on the so far largest variety of interaction points (several of
which are consumer-initiated interactions), Part V does so by explicitly focusing
on the consumers point of view throughout multiple phases of the buying process.
Additionally, it is worthwhile to note that by adopting the status quo bias theory,
the dissertations multidisciplinary perspective is reflected.
2006). Thus, besides the theoretical contributions, the methodology was adapted
to fit the purpose and the state of the theory (Edmondson and McManus, 2007).
Although value has been accepted as the first and foremost driver of exchange,
only initial ideas regarding the impact of the service- and the customer-dominant
logic have been published. To fill this gap in literature, this article discusses
the impact of both logics on the customer value concept. A conceptual model
focusing on interactions with customers and the integration of the customer into
the companies processes is derived. The article has been accepted for publication
in the "Der Markt: International Journal of Marketing".
The second article directly applies the ideas of the service-dominant logic to
the creation of a service brand. In particular, the spotlight is put on the perceived
employer brand and its role in the creation of a service brand by employee-
customer interactions. Doing so, the article investigates employee satisfaction and
identification with the company as antecedent factors to the creation of a service
brand. Most of the hypothesized relationships between the perceived employer
brand and favorable outcomes were confirmed. The article is already published in
the "Journal of Services Marketing".
Article 3: Accessing the Biggest Piece of the Pie: International Segmentation with
Emerging Markets
Conference 2012" and the "41st Annual Conference of the European Marketing
Association". At the latter, it was nominated for the best paper award based on a
doctoral dissertation. Further, the article has yet passed the first review round in
the "Journal of International Marketing".
Article 4: Nobody said Leaving was Easy: Biased Consumer Behavior in Research
Shopping
Focusing on the consumers point of view in the buying process, the fourth
article investigates research shopping, which is delineated as the tendency to
switch channels between search and purchase phase. Applying the status quo
bias theory, the following question is investigated: "Why are consumers reluctant
to research shopping"? The findings of two studies, a large-scale survey and
a laboratory experiment, suggest that consumers who use a single channel for
searching, and in particular consumers satisfied with the search channel, are
prone to a status quo bias and therefore, reluctant to research shopping. The
article is in preparation for submission to the "International Journal of Research
in Marketing".
B IBLIOGRAPHY 11
Bibliography
Rynes, S. and R. Gephart Jr, From the editors. Academy of Management Journal,
47(4):454-462.
TechCrunch. 2011. Its still a feature phone world: Global smartphone penetration
at 27%. (Available at www.techcrunch.com).
Vargo, S. L., P. P. Maglio, and M. A. Akaka, 2008. On value and value co-creation:
A service systems and service logic perspective. European Management
Journal, 26(3):145-152.
Yahoo! Inc. and OMD, 2006. Long & winding road: The route to the cash register.
(Available at http://files.shareholder.com/).
13
Abstract
Customer Value (CV) is one of the most crucial concepts in the field of marketing.
Literature states that the identification and creation of CV is decisive for the
strategic success of any organization. Moreover, CV was coined a hot research
topic in the field of marketing for the years of 2010-2012 by the Marketing Science
Institute (2010). However, there is still no clear opinion about the concept per
se, as there is currently no accepted CV definition. What is more, since a broad
variety of fundamental developments, such as the rise of the service-dominant and
customer-dominant logic, touched the CV concept in the past years, there is a need
to reflect upon them. Therefore, the article addresses this challenge and analyzes
recent developments to provide a common basis for future research. Moreover,
the resulting implications for CV creation are provided. Focal questions that arise
when considering CV include: What is the nature of CV and how can value be
created considering the implications of both logics?1
Keywords: Customer value, S-D logic, C-D logic, Nature of value, Value creation
1 T. Schlager and P. Maas. This paper has been accepted for publication in the "Der Markt:
1 Introduction
Within the past few decades, there has been a broad shift from searching for
sources of competitive advantage within a company, to investigating external
sources of competitive advantage (Cleland and Bruno, 1996; Khalifa, 2004;
Woodruff, 1997). One of marketings most fundamental concepts, customer value
(CV), recognizes this by placing the focus on the customer. The relevance of
CV is not under discussion, as the creation of CV is considered the basis for a
companys business success (Huber et al., 2001; McDougall and Levesque, 2000;
Payne and Holt, 2001; Porter, 1996; Woodruff, 1997), central to competitive
advantage (Khalifa, 2004; Kothandaraman and Wilson, 2001; Parasuraman, 1997),
and directly related to shareholder value (Cleland and Bruno, 1996). In 2007, the
American Marketing Association (2007) adapted its definition of marketing to
reflect the focus on value. More recently, Vargo and Lusch (2008a) defined value
as "idiosyncratic, experiential, contextual and meaning laden" (p. 7). However,
due to its complexity, there is still much discussion about the meaning of CV,
as several definitions have been attached to the term (Smith and Colgate, 2007).
The semantic heterogeneity is also highlighted by the multitude of different terms
that have been used in the context of CV (Voima et al., 2010). Recently, the
term customer has even come under discussion (Henneberg and Mouzas, 2008;
Vargo and Lusch, 2004; Woodruff and Flint, 2006). Consequently, the marketing
literature is still deficient in its understanding of CV (e.g., Blocker, 2011; Smith
and Colgate, 2007; Woodruff and Flint, 2006) and calls for further investigation
into CV are loudening (Voima et al., 2010).
the C-D logic, acknowledge the importance of the customer, and thus, inherently
strengthen the focus on the customer and, as a result, the CV concept. On the
other hand, the logics comprehensively summarize issues changing and advancing
CV (Voima et al., 2010; Woodruff and Flint, 2006). We regard the impact on CV
as twofold. Clearly, the scope in which CV is considered has risen (Edvardsson
et al., 2011; Gummesson, 2008; Woodruff and Flint, 2006), which makes CV
a concept that is even more vague than before. However, one might also argue
that the mentioned logics have a large potential to inform, and thus, to further
develop the CV concept (e.g., Voima et al., 2010; Woodruff and Flint, 2006). As
we believe more strongly in the second aspect, we acknowledge the importance of
registering the impacts for further enhancing CV.
The primary objective of our paper is to fill this gap in the literature and to
provide a basis for further research by evaluating the influences of the S-D and
C-D logic (Heinonen et al., 2010; Vargo and Lusch, 2008a; Vargo and Lusch,
2004). In particular, our article addresses the following research questions: How
did both logics change the understanding of CV? As implications regarding the
creation of value can primarily be recognized by taking into account the S-D and
16 II D OMINANT L OGICS
C-D logics perspectives, we also discuss what this means for value creation. In
doing so, we make particularly three contributions to the literature on CV:
The nature of CV
The scope of this paper is limited to the nature and the creation of CV, as our
primary objective is to provide a starting point for other scholars. Neither do
we enter in the current discussion, like the one regarding the roles within value
creation (e.g., Grnroos, 2011), nor do we aim to provide a completely new
conceptualization regarding CV, as we only focus on the changes implied by the
S-D and C-D logic.
The remainder of the article is organized as follows (see Figure 3). First, the
S-D and C-D logic are outlined with a particular focus on the resulting changes in
current value paradigms (i.e., from value-in-exchange to value-in-use). Next, we
discuss the nature of CV and the range of CV considerations as major changes
and enhancements are proposed by both logics. As a primary focus of the S-D
and C-D logic is laid on value creation, we outline ways of creating CV from
this novel perspective in a third step. Finally, a model of CV, from the logics
perspective, is established and briefly delineated by summarizing the key points.
While each section of the article emphasizes the current state of the literature, we
broadly outline avenues for future research, as well as theoretical and practical
implications, as we strive to provide a common parting ground.
2 The dominant logics and value concepts 17
The S-D logic was initiated in 2004 by the award-winning article Evolving to a
new dominant logic for marketing (Vargo and Lusch, 2004). It accentuates a shift
away from goods to service provision, lays a focus on intangible resources, the co-
creation of value by the customer, and the notion of interaction and relationships,
and proposes a phenomenological facet of value (e.g., Vargo and Lusch, 2006).
Woodruff and Flint (2006) back the importance of the S-D logic for CV by
2 When using the logics, we refer to both upcoming logics as in a multitude of aspects, as they
appear to have similar consequences. In case they differ, we only refer to the applying logic.
18 II D OMINANT L OGICS
Recently, the C-D logic appeared in the literature picking up thoughts of the
S-D logic, as well as passing criticism on the S-D logic: According to the C-D
logic, the S-D logic is still a supplier-dominant logic. Arguing so, the C-D logic
even goes further in positioning the customer in the center by shifting the focus
from the companys processes to the customers reality and history (Heinonen
et al., 2010). The primary distinction between both logics is the point of view.
While the S-D logic is supplier focused, the C-D logic refers to the customer as the
starting point of most relevant considerations. For instance, under the C-D logic,
service providers need to enhance their understanding of the customers history,
objectives and goals, followed by using services (Voima et al., 2010; Wikner,
2010). From a CV perspective, we regard this inherent focus on the customer as
useful, even though the C-D logic has primarily been applied to a service context.
Thus, although only a handful of articles have appeared regarding this logic, we
apply its notions on CV.
Both previously-mentioned logics, especially the rise of the S-D logic, have
a fundamental impact on the development of CV (Voima et al., 2010; Woodruff
and Flint, 2006). What they have in common is that they both emphasize the
importance of the CV concept and impose foundational challenges for the CV
concept (Heinonen et al., 2010; Vargo and Lusch, 2008a). More specifically, they
enhance discussions about value considerations that are closely tied to the nature
of CV and thus, appear as highly relevant for the presented article.
As both logics influence the nature of CV, which has already been vague before,
there is a need to address them. Up to now, several facets of CV have repeat-
edly appeared in the literature regarding CV. Among those, CV is described as
being a trade-off between benefits and sacrifices (Ulaga, 2003; Zeithaml, 1988;
Gale, 1994), subjective (Kortge and Okonkwo, 1993; Nasution and Mavondo,
2008; Ulaga, 2003; Vargo and Lusch, 2004), multi-dimensional (Holbrook, 2005;
Matzler, 2000; Sweeney and Soutar, 2001; Ulaga, 2003) and relativistic (Gale,
1994; Holbrook, 1994, 2006). Moreover, it has been delineated as context-specific
and experiential (Holbrook, 2005). Although the ideas of several facets have
been picked up by both logics (i.e., CVs context-specific and experiential nature),
substantial changes can be recognized which are summarized in the following
section. Figure 4 presents the development from the classical understanding of
CV to an understanding from a S-D and C-D logics perspective; each facet is
assigned the logic that mostly emphasizes its notion.
According to the S-D logic, a shift towards emphasizing the experiential and
phenomenological facet of CV can be acknowledged by the shift towards em-
phasizing the experiential and phenomenological facet of CV (Vargo and Lusch,
3.1 Phenomenological and experiential facet 21
2004; Woodruff and Flint, 2006). Both terms, experiential and phenomenological,
emphasize the co-creation role of the customer (Vargo and Lusch, 2008a). From
this understanding, experience does not refer to the customers past experiences;
it rather describes the perishability inherent to CV (Vargo and Lusch, 2008b).
This underscores that value cannot be inventoried and is not created solely by the
company that is, however, not new to the literature.
Prahalad and Ramaswamy (2004b, p. 137) use the term in a similar man-
ner and emphasize the shift toward experiences as they state that "value is now
centered in the experiences of consumers". Similarly, even earlier contributions
suggest an experiential perspective that includes symbolic, hedonic, and aesthetic
aspects in the consumption process (Babin et al., 1994; Hirschmann and Hol-
brook, 1982). Although both terms are still discussed, phenomenological has
recently been preferred, as experiential implies several other meanings, such as
a focus on the past. In contrast, phenomenological emphasizes the idiosyncratic
determination of value (Vargo and Lusch, 2008a) without implying a focus on
the past. This accentuates that companies and customers have equal roles in
value determination, as the process between both parties is emphasized, while
the old-fashioned notion of value being provided by the company is no longer
supported. The idea is explicitly expressed by highlighting interactions when
considering the creation of CV as the S-D logic does. Woodruff and Flint (2006)
explain the phenomenological nature of CV as context-specific, interactive, and
attached with meanings.3 The C-D logic accepts that interactions facilitate the
creation of value, however, contrary to the S-D logics supplier-oriented approach,
it emphasizes that it is the customer, who ultimately determines the value created.
In doing so, the C-D logic stresses that other processes not directly related to
interactions also need to be considered (Heinonen et al., 2010).
context specificity, we treat this as a separate issue, as it has not been clearly delineated by the authors.
Moreover, a multitude of other articles has further developed the meaning of context-specific without
referring to it as the phenomenological aspect. Hence, we discuss this issue separately.
22 II D OMINANT L OGICS
While the previous literature has accepted CVs dependence on the context,
the S-D and C-D logic have strongly emphasized and further developed this
idea. One of the first guiding conceptual articles was put forth by Edvardsson
et al. (2011). Social construction theories are applied to stress the importance
of the social context for value creation under the S-D logic. On this basis, four
propositions are delineated, two of which strongly underline the importance of the
social dimension of value: "Value has a collective and intersubjective dimension
and should be understood as value-in-social-context" (p. 333) and "the way in
3.3 Experience-based facet 23
which resources are assessed depends on the social context" (p. 334). It can be
concluded that research efforts begin to adapt a more differentiated point of view,
which includes the social context as an important variable for the determination of
value.
The C-D logic argues in a similar manner, stating that the dyadic approach that
was historically used is not enough to display the complex construct of CV (e.g.,
Voima et al., 2010). Hence, this logic also highlights that the customer is socially
embedded, interacting with other groups, such as other customers. Compared to
the S-D logic, the C-D logic more strongly emphasizes the customers point of
view. An even stronger focus is laid on the customer, and thus, on his/ her specific
context. In sum, while the social context is emphasized by both logics, the C-D
logic fully takes in the customers perspective (Voima et al., 2010).
The S-D and C-D logic emphasize a fundamentally changed perspective on value
and value creation. Intuitively resulting from discussions regarding CVs nature is
the question: How is CV creation considering this new standpoint? We address
this question by recognizing that former frameworks (i.e., the one put forth by
Smith and Colgate, 2007) have yet become outdated in the light of S-D and C-D
logic. What is worthwhile to note is that discussions about the role of the company,
as well as the customer in value creation, have appeared (Edvardsson et al., 2011;
Grnroos and Ravald, 2011; Gummesson, 2008). We do not directly enter these,
but instead reflect implications from the changing nature on a general level for
two reasons. First, by drawing implications on value creation on a higher level,
we aim to provide a basis for enhancing this discussion, while entering these
would go beyond the scope of this paper. Second, we regard discussions as highly
dependent on the point of view and the context (i.e., product vs. service context,
4.1 Interaction 25
Gummesson, 2008), which we do not discuss either. Thus, we take this avenue
and address questions, which are more general in nature. More specifically, we
identified interactions, a focus on relationships, and holistic marketing processes
as responsible for value creation.
4.1 Interaction
The recent literature concerning the S-D logic and value-in-use places a special
emphasis on company-customer interactions as source of CV. Although interac-
tions have already been acknowledged as source of value by other frameworks
(e.g., Smith and Colgate, 2007), the notion is further enhanced. Emphasizing
interactions as a crucial concept from a S-D logic perspective, Grnroos and
Ravald (2011, p. 12) define interactions as "a mutual or reciprocal action where
two or more parties have an effect upon each other". On this basis, it is argued that
through interactive processes, companies can get actively involved in creating ex-
periential value. Other contributions confirm this by seeing interactions as having
the ability to promote experiential and phenomenological value (Ballantyne and
Varey, 2006). Similarly, scholars assess that interactions provide the basis for form-
ing the customers preferences (Mathwick et al., 2002). The importance of the
company-customer interaction is underlined by the ability to facilitate value and
to influence CV perceptions. As a result, it is broadly accepted that the customers
creation of value is catalyzed through interactions (Grnroos, 2009; Grnroos
and Ravald, 2009, 2011; Tynan et al., 2010). Although the difference between
services and goods environments is still discussed, it appears that interactions
being especially important in a service-context is accepted. From the companies
perspective, this is due to the multitude of opportunities for co-creating value with
the customer when acting within the customers sphere (Grnroos, 2008).
One other issue that comes into play is the superior ability to sense the cus-
tomers needs within interactions (Tynan et al., 2010). Although interactions per
se are especially considered by the S-D logic, they may also provide a source of in-
26 II D OMINANT L OGICS
depth knowledge about the customer and his/her life, which is rather emphasized
by the C-D logic. More information can be generated, and as a result, companies
can develop better insights about their customers. This issues importance is
underscored by the C-D logic; therefore, we also argue that from a C-D logic
point of view those company-customer interactions are highly relevant.
4.2 Integration
Closely connected to, but still distinct from interactions as source of CV, is the
integration of resources (e.g., Mele et al., 2010). This includes the integration of
the customers resources which refers to providing customers with the opportunity
to participate within the companies processes (e.g., Mller, 2006). Integrating the
resources of customers and companies is treated as prerequisite for interactions.
As Grnroos and Ravald (2011, p. 11) state "an inherent aspect of interaction is
connectivity, i.e., the parties involved are in some contact with each other". With
the term connectivity, the authors emphasize the integration of the companies and
the customers resources. The importance of the integration of the companies
and customers resources in the value creation process is strongly emphasized by
the S-D logic (e.g., Vargo and Lusch, 2004). In a similar manner, Edvardsson et
al. (2010) refer to getting close to customers by customer integration.
We argue that this proximity to the customer partly takes in an avenue that
is emphasized by the C-D logic. The C-D logic accepts the importance of in-
teractions, but even goes one step further by also seeing "other activities and
experiences" (Heinonen et al., 2010, p. 535) that are not directly connected to
service-related interactions. From this position, customer processes, which are
not controlled by a company, are considered as a highly relevant part within
value creation. Hence, the challenge that needs to be addressed is discovering
the underlying issues that cannot be easily recognized by a company. When the
customers life is brought into the focus of value creation, an in-depth knowledge
of the customer becomes necessary. As Heinonen et al. (2010, p. 538) note, "we
4.3 Focus on relationships 27
first need to understand the customers lives" which renders the importance of a
fine-grained understanding of the customer. One way to gain this depth of under-
standing of the customer is integrating customers into the companys processes.
The aim of integration may therefore not only be the creation of interactions,
but also the achievement of profound customer insights that go beyond insights
gained by traditional marketing research methods. Hence, integration can also be
considered a way for getting deep customer insights regarding what the C-D logic
describes as value-in-life.
When facing notions like interaction and integration, it becomes obvious that
relationships play an important role within value creation, which is a concept
that is extensively highlighted by both logics (Vargo and Lusch, 2004; Voima
et al., 2010; Woodruff and Flint, 2006). Both interaction and integration can
be considered a means to the end, which is relational value. This is due to
the important role of interactions in facilitating relationships (Payne et al., 2008;
Prahalad and Ramaswamy, 2004a; Sandstrm et al., 2008; Vargo and Lusch, 2004).
Generally, it is proposed that relationships themselves are a source of value (e.g.,
Howden and Pressey, 2008; Lindgreen and Wynstra, 2005); relational benefits
and sacrifices need to be considered in the value formation process (Grnroos,
1997; Payne and Holt, 2001; Ravald and Grnroos, 1996; Ulaga and Eggert, 2005).
In this light, trust and risk avoidance are especially considered focal constructs
(e.g., Lindgreen and Wynstra, 2005). Most commonly, the literature assesses the
value that emerges in relationships between companies and customers (Ulaga and
28 II D OMINANT L OGICS
Eggert, 2005, 2006b), but also the relationship between customers and customers
(Edvardsson et al., 2011; Voima et al., 2010) and between customers and brands
(Payne et al., 2009).
Similarly to the explanation regarding the integration of the customer into the
companys processes, we argue about the focus on relationships. Relationships are
a source of value, but they also provide companies with a good understanding of
the customer and his/her needs. A multitude of studies has connected relationships
with the customers willingness to provide information (e.g., Lapierre, 2000;
Wuyts and Geyskens, 2005). Even intimate information regarding the customers
life can be sensed within a close relationship (Grayson, 2007; Wong et al., 2005).
Hence, information that is more accurate can be gained concerning what customers
value considering their life and experiences. Therefore, we propose that the
relevance of relationships can also be considered as an important part in fitting the
companies actions and offers to the customers life, stressed by the C-D logic.
In summary, although the relational facet is not new to the literature, the focus
on interaction and integration augments the importance of relationships from
both logics points of view. Thus, focusing on relationships can be proposed as a
fruitful way of creating value in a S-D and C-D logic context.
The customers relationships to other actors, and thus, the notion of customers
being embedded within a highly complex context, is emphasized by the two
discussed logics. Again, one catalyst is the rise of the S-D logic that emphasizes
the customer as resource integrator also using other resources for value creation
(e.g., Vargo and Lusch, 2004). Consistent with this, developments, such as shift to
the empowered and active customers (Edvardsson et al., 2011; Payne et al., 2009;
Vargo and Lusch, 2004), customer-customer and social interactions, have rapidly
gained importance. A similar argument holds true from a C-D logic point of
view, which emphasizes processes in value creation that are not directly related to
4.4 Holistic marketing processes 29
services (Heinonen et al., 2009). Therefore, we briefly review processes that have
been discussed for creating value in the face of the complexity delineated by the
S-D and C-D logic perspectives. In this light, Edvardsson et al. (2011) emphasized
the customers context as a source of value for the customer. That is, appearing
more valuable to the social context enhances the individuals value perception.
Hence, the relevant question to ask becomes: How can value creation through the
social context be enhanced? We argue that an initial answer to this question can
be found in the already discovered constructs and propose that branding provides
a fruitful way to enhance value creation within a complex social context, as we
will outline below.
Brodie et al. (2009) summarized previous articles about the influence of brand-
related constructs on perceived customer quality and perceived value. Similarly,
other articles have discussed the relationships between CV and branding (Brodie
et al., 2009; Nasution and Mavondo, 2008; Payne et al., 2009; Tynan et al., 2010).
A positive brand image indirectly enhances value-in-use (Strandvik and Rindell,
2010), a development which can be considered as one of the assumptions of the
S-D and C-D logic. Similarly, the social context is emphasized by both logics;
consumers form relationships with brands that reflect their social relationships
(Payne et al., 2009), and thus, reflect the customers accumulated experiences and
knowledge. Brands therefore act as comprehensive facilitator of CV. Although
earlier contributions have argued that resources, like brands, may constitute a
source of value (e.g., Dawar and Parker, 1994; Erdem and Swait, 1998; Mller,
2001; Srivastava et al., 2001), the idea of relating branding and company images
to CV is a more recent phenomenon and has become increasingly popular. For
instance, the importance of experiential value is emphasized by both logics (Vargo
and Lusch, 2004; Voima et al., 2010). Cova and Pace (2006) closely relate the
research on communities to branding by mentioning brand communities and
brand tribes. Accordingly, brand communities and tribes may leverage a brands
symbolic value, and therefore, provide a source of CV. Consistent with this Cretu
and Brodie (2007) confirm through their research that official and unofficial brand
30 II D OMINANT L OGICS
Following the aim of our paper, which is to register influences of dominant logics
on CV, on the one hand, we addressed the nature of CV. On the other hand, we
mirrored the nature in terms of implications for CV creation. The model shown in
Figure 5 underpins the papers idea, which is to draw recommendations regarding
value creation based on the nature of CV. Hence we connect both sides, the
nature of value with the respective source of value creation. As we have already
comprehensively explained both issues, we outline the model on a high level.
about value are shown, which form the basis for explicating CV and relating it to
value creation. The increasing complexity of CV considerations is shown first as
value-in-use, then value-in-context, and finally value-in-life. Value-in-life and the
experience-based facet underpin the importance of in-depth knowledge, which is
displayed by insights resulting from interactions, customer integration and close
relationships. Regarding the value-in-life, practical approaches still need to be
found, as currently no research is dedicated to enhancing this type of value.
Certainly, the model remains abstract and does not display all of the complex
processes within the area of CV and the ways of creating CV, however, we argue
that the most relevant implications of our manuscript, and thus, of both logics, are
condensed to a model.
6 Discussion
Until today, no single opinion has been established regarding CV (Gallarza and Gil
Saura, 2006; Graf and Maas, 2008; Smith and Colgate, 2007). Within this vague
6 Discussion 33
context of CV, we recognize the ambiguous influence of both logics that changes
marketing literature. On the one hand, the S-D and C-D logic add complexity
to CV, making it an even more unseizable concept, while on the other hand, the
logics enrich the understanding of the concept as CV is inherently compatible
with the ideas of both logics (Voima et al., 2010; Woodruff and Flint, 2006). As
we believe in the second issue, we reframed the considerations about CV from
the perspective of currently discussed logics. We argue that by doing so, we
provide more clarity for the concept of CV. For instance, our article addresses
the embeddedness of both, the customer and the company, in the light of value
creation. Moreover, the novelty and relatedness of both logics to the value concept
strongly emphasizes the relevance of our paper: There is a clear need to reflect
their impacts on the CV literature.
One might argue about including the C-D logic as only few contributions have
been made so far. However, we see two important reasons for doing so. Firstly,
its thoughts are not completely new to the literature, as the considerations of the
C-D logic build upon well-established concepts (i.e., the inherent customer focus,
Drucker, 1974 cf. Heinonen et al., 2010). Secondly, as previously outlined, we
comply with articles concerning the C-D logic by arguing that its ideas further
refine the S-D logic with a focus on the customer (Heinonen et al., 2010; Voima
et al., 2010) and thus, have the potential to further inform the CV literature. In
this regard, the opinion of researchers arguing in favor of a C-D logic state that
the S-D logic is still deficient in several points (Heinonen et al., 2009); these are
included in our manuscript.
An issue that is still under discussion in the value literature is the question of
who is the value creator at all (e.g., Grnroos, 2008; Gummesson, 2008; Lusch
and Vargo, 2006a). On purpose, we did not enter this discussion because the
objective of our paper was to inform the CV literature with what is well-known
to date and would have exceeded the scope of the presented article. Hence, we
state that who creates value still needs to be reviewed by a multitude of articles.
34 II D OMINANT L OGICS
from S-D and C-D logic might lead to different conclusions in some areas. The
primary distinction is the C-D logics focus on the customers point of view,
while the S-D logic argues from a supplier perspective (e.g., Voima et al., 2010).
Thus, articles regarding CV might come to a different conclusion based on the
perspective applied. Researchers shall thoroughly evaluate whether to argue from
a S-D logic or a C-D logic point of view. For instance, when focusing on the
customers past in determining CV, one may be better advised to apply a C-D
logic, rather than an S-D logic approach. Whether the literature will reach a
consensus is still to be determined by future articles, however, we believe that this
article provides a solid basis to start from.
8 Future research
Clearly, many articles have discussed the nature and creation of CV within the
last decade. Nevertheless, further research is needed to enhance the understanding
of the concept. For instance, the phenomenological nature of CV still calls
for further research, as discussions around the S-D logic are still ongoing (e.g.,
Mele et al., 2010). For example, the influence of companies is questioned when
value is phenomenological in nature (Heinonen et al., 2009). As these valid
issues arise, it becomes obvious that it still remains unclear what "uniquely and
phenomenologically determined by the customer" (Vargo and Lusch, 2008a, p.
7) means. Thus, creating value from a phenomenological perspective needs to
be investigated more extensively. However, the importance of use-situations
has broadly been overlooked by research on CV. Therefore, we suggest that
considering co-creation opportunities may enrich the current literature of CV
in order to better express the notion of CV. In line with this argument is that
the literature on the creation of CV has not profoundly evaluated the role of
customer integration and interaction, most commonly, only specific issues have
8 Future research 37
been assessed (e.g., Mele et al., 2010). Initial research may evaluate the influence
that providing opportunities for interaction have on the evaluations of CV. From
this starting point, scholars might enhance research on this issue and approach
more detailed research questions.
are invisible. Clearly, deep customer insights need to be accumulated for doing
so. We propose further integrating the customer into the companys processes,
and hence, determining his/her desires in case they change. Similarly, how to
deepen the relationships to customers might be a fruitful way for future research.
Nevertheless, no research has appeared regarding this issue due to its novelty.
Thus, we suggest that tackling this issue is a fruitful and informing approach.
Bibliography
Babin, B. J., W. R. Darden, and M. Griffin, 1994. Work and/or fun: Measur-
ing hedonic and utilitarian shopping value. Journal of Consumer Research,
20(4):644-656.
Brodie, R. J., 2009. From goods to service branding. Marketing Theory, 9(1):107-
111.
B IBLIOGRAPHY 41
Cova, B. and R. Salle, 2008a. Creating superior value through network offer-
ings. Creating and managing superior customer value. In A. G. Woodside,
F. Golfetto, and M. Gibbert, editors, Advances in business marketing and
purchasing, pages 317-342, Emerald Group Publishing Limited. Bingley, UK.
Cretu, A. E. and R. J. Brodie, 2007. The influence of brand image and company
reputation where manufacturers market to small firms: A customer value
perspective. Industrial Marketing Management, 36(2):230-240.
Gale, B. T. 1994. Managing customer value: Creating quality and service that
customers can see. The Free Press, New York, NY.
Gallarza, M. G. and I. Gil Saura, 2006. Value dimensions, perceived value, satis-
faction and loyalty: An investigation of university students travel behaviour.
Tourism Management, 27(3):437-452.
Grieve, K. and E. Ortiz, 2003. Customer value management sins can be costly.
DM Review, 13(11):38-40.
Grnroos, C., 2008. Service logic revisited: Who creates value? And who
co-creates? European Business Review, 20(4):298-314.
Grnroos, C., 2011. A service perspective on business relationships: The value cre-
ation, interaction and marketing interface. Industrial Marketing Management,
40(2):240-247.
Grnroos, C. and A. Ravald, 2009. Marketing and the logic of service: Value
facilitation, value creation and co-creation, and their marketing implications.
Working Paper, No. 542, Hanken School of Economics, Helsinki, Finland.
(Available at helda.helsinki.fi/).
Levitt, T, 1986. The marketing imagination. The Free Press, New York, NY.
Lusch, R. F., S. L. Vargo, and M. Tanniru, 2010. Service, value networks and
learning. Journal of the Academy of Marketing Science, 38(1):19-31.
Mathwick, C., N. K. Malhotra, and E. Rigdon, 2002. The effect of dynamic retail
experiences on experiential perceptions of value: An internet and catalog
comparison. Journal of Retailing, 78(1):51-60.
Merz, M. A., Y. He, and S. L. Vargo, 2009. The evolving brand logic: A service-
dominant logic perspective. Journal of the Academy of Marketing Science,
37(3):328-344.
Payne, A. and S. Holt, 2001. Diagnosing customer value: Integrating the value
process and relationship marketing. British Journal of Management, 12(2):159-
182.
Payne, A. F., K. Storbacka, and P. Frow, 2008. Managing the co-creation of value.
Journal of the Academy of Marketing Science, 36(1):83-96.
Payne, A., K. Storbacka, P. Frow, and S. Knox, 2009. Co-creating brands: Di-
agnosing and designing the relationship experience. Journal of Business
Research, 62(3):379-389.
Ravald, A. and C. Grnroos, 1996. The value concept and relationship marketing.
European Journal of Marketing, 30(2):19-30.
Shah, D., R. T. Rust, A. Parasuraman, R. Staelin, and G. S. Day, 2006. The path
to customer centricity. Journal of Service Research, 9(2):113-125.
Tynan, C., S. McKechnie, and C. Chhuon, 2010. Co-creating value for luxury
brands. Journal of Business Research, 63(11):1156-1163.
Vandermerwe, S., 1996. New competitive spaces: Jointly investing in new cus-
tomer logic. Columbia Journal of World Business, 31(4):80-101.
Vargo, S. L., P. P. Maglio, and M. A. Akaka, 2008. On value and value co-creation:
A service systems and service logic perspective. European Management
Journal, 26(3):145-152.
Voima, P., K. Heinonen, and T. Strandvik, 2010. Exploring customer value for-
mation: A customer dominant logic perspective. Working Paper, No. 552, Han-
ken School of Economics, Helsinki, Finland. (Available at https://helda.helsinki.fi/).
Woodruff, R. B., 1997. Customer value: The next source for competitive advan-
tage. Journal of the Academy of Marketing Science, 25(2):139-153.
Zhou, K. Z., J. J. Li, N. Zhou, and C. Su, 2008. Market orientation, job satisfaction,
product quality, and firm performance: Evidence from China. Strategic
Management Journal, 29(9):985-1000.
52 III S ERVICE B RANDING
Abstract
4 T. Schlager, M. Bodderas, P. Maas, and J. L. Cachelin. This paper has been presented at the
11th International Research Conference in Service Management and the 30th annual meeting of the
Strategic Management Society". It is published in the "Journal of Services Marketing".
1 Introduction 53
1 Introduction
To the best of our knowledge, the concept of the EB has not yet been proven
54 III S ERVICE B RANDING
as relevant for SB. We believe that our idea has two major advantages. On the one
hand, employee-customer interactions can be managed more efficiently than other
interactions that involve the customer such as customer-customer interactions,
which can rarely be directly influenced. On the other hand, the concept of the EB
is especially relevant in knowledge intensive contexts (Berthon et al., 2005). This
knowledge intensity is further emphasised by a service-centred context (Vargo and
Lusch, 2008). Recognising this potential, the central objective of our approach
is to make progress in relating the concept of the EB to SB. For this purpose,
we empirically examine the influence of EB dimensions that positively affect
employee attitudes, which in turn have been conceptually and empirically proven
to have a positive impact on customer experiences.
The rest of the article is structured as follows: First, we review branding within
a service-centred context and outline the importance of employees. Next, we
discuss the concept of the EB and why we propose it as an approach relevant to
SB, thus accounting for our research framework. Third, we present the hypotheses
that link the EB to SB-relevant employee outcomes. Finally, we empirically test
these hypotheses and conclude with a discussion.
2 Conceptual development
For the development of our conceptual model (Figure 6), we review the literature
on branding, SB, the service profit chain and employer attractiveness. Based on
this, we concretise the constructs of our framework and establish the hypotheses.
concept and employer branding. The first definition of the EB was given by
Ambler and Barrow (1996, p. 187), who defined the concept as "the package
of functional, economic and psychological benefits provided by employment
and identified with the employing company". In this approach to the EB, the
primary objective is to "provide a coherent framework for management to simplify
and focus priorities, increase productivity and improve recruitment, retention
and commitment". Thus, the main objective of the EB concept is to positively
influence current and prospective employees in order to attract high-potential
employees (Berthon et al., 2005).
In order to set up a framework that comprises both the EB and SB, we build
upon the research of Brodie et al. (2009). They further developed a framework
on SB that was initially set up by Calonius (1986) and refined by other scholars
(Bitner, 1995; Grnroos, 1996, 2006, 2007). Their framework on SB discusses
2 Conceptual development 57
The framework of Brodie et al. (2009) argues that a company should first
create the service brand through internal marketing between the company and the
employees. A strong EB contributes to triggering favourable employee attitudes
(Berthon et al., 2005). This, in turn, leads to the second process of SB: The inter-
active marketing perspective. Here, employees are the main driver for customer
experiences and thus, for SB (Brodie et al., 2009; Hatch and Schultz, 2001). Con-
sequently, it can be assumed that the EB fosters the attitudes of customer-contact
employees, which directly influence customer attitudes toward the brand (Mosley,
2007; Papasolomou and Vrontis, 2006). Moreover, the potential employees per-
ceptions of the brand are reinforced as they sense the current employees positive
attitudes (Gatewood et al., 1993). Finally, SB includes the external marketing
perspective, which includes all processes of direct communication with customers.
Considering the external marketing perspective, we only empirically investigate
the direct positive effect on potential employees, and not the effect on customers.
Nevertheless, potential employees may be an important future link to consistent
SB, a point which emphasises the relevance of investigating this factor. The re-
sults of the discussion of the three relationships support that forming a strong EB
contributes to establishing a consistent service brand image. Figure 6 summarises
the proposed framework that links SB and the EB.
58 III S ERVICE B RANDING
3 Hypotheses development
As the aim of our study is to research employee attitudes that provoke favourable
customer attitudes towards a brand, we must first determine relevant customer
attitudes. The first customer outcome that may predict future behaviour towards a
brand (Mittal and Kamakura, 2001) and provokes positive brand attitudes (Spreng
et al., 1995) is customer satisfaction. Furthermore, it has a significant effect
on attitudes as well as on loyalty toward a brand (Spreng et al., 1996), which
underpins its relationship to the creation of the service brand. Second, the ultimate
goal of the brand-building process is to create a brand relationship (Aaker, 1991;
Keller, 1993), which is reflected by the level of a customers identification with the
brand (Kimpakorn and Tocquer, 2010). Furthermore, a customers identification
with a company distinguishes the brand from others, which is one of the brands
most important aims (Kim et al., 2001).
On a high level, scholars divide the concept of the EB into two parts. On the
one hand, researchers take into account general attitudes towards the company. On
the other hand, perceived job attributes are included when constructing measure-
ments for the EB (Berthon et al., 2005; Collins and Stevens, 2002). In developing
a measurement instrument for the EB, we partly applied the items put forth by
Berthon et al. (2005) that measure the perceived EB for potential employees. We
included three of their five dimensions: Economic value, development value, and
social value, as they seemed best suited to predict favourable employee attitudes
in the study by Berthon et al. (2005). However, we added two other dimensions:
Diversity value and reputation value, which are identified as important for em-
ployee attitudes as will be outlined in the following. Furthermore, we slightly
revised the items of the dimensions of Berthon et al. (2005) in order to better
reflect the investigation of current employees.
The third dimension, social value, encompasses a strong team spirit, competent co-
workers, a friendly relationship amongst individual co-workers, and a respectful
environment. The items that constitute social value have frequently been assessed
by scholars as they appear to be important drivers for positive employee attitudes
(e.g., Saari and Judge, 2004). The role of co-workers and supervisors (competent
or friendly) in enhancing favourable employee attitudes like satisfaction has
been well-established in literature (e.g., Saari and Judge, 2004). Satisfaction
has a positive impact on employee commitment, a construct strongly related to
employee identification with the company (Avery et al., 2007; Harrison et al.,
2006). In investigating the role of employee satisfaction as a moderator between
62 III S ERVICE B RANDING
social value and employee identification, there is evidence that there is indeed
a direct relationship between social value and employee identification. Social
value has been investigated in numerous fields such as literature concerning
organizational justice, social exchange, or organizational citizenship behaviour.
The idea common to all three fields is that employees behaviour and attitudes can
arise toward different groups within an organization, such as whole departments or
individual co-workers (Lavelle et al., 2007). Organizational justice, for example,
creates commitment and identification with the organization.
The fourth dimension, diversity value, refers to interesting job characteristics and
is similar to interest value conceptualised by Berthon et al. (2005). However, we
added several items that have been mentioned in the context of the EB such as
challenging tasks (Backhaus and Tikoo, 2004; Saari and Judge, 2004; Towers
Perrin, 2005) and a broad variety of tasks (Backhaus and Tikoo, 2004). Studies
have shown that interesting job characteristics, and thus diversity value, are
one of the most important determinants of job satisfaction (Judge and Church,
3.5 Reputation value 63
2000; Jurgensen, 1978), and others assessed it as the best predictor of overall job
satisfaction (Saari and Judge, 2004). The amount of interesting work is found
to influence the level of organizational identification (Lee, 1971). It is crucial
to realise that potential employees may easily perceive these characteristics of a
company as functional aspects, challenges, or unique opportunities of employment
during the recruiting process and job descriptions provided by companies. This
leads to the following hypotheses:
reputation for attracting new high-potential employees (Cable and Graham, 2000;
Preston and OBannon, 1997).
4 Empirical study
An insurance provider was chosen as the object of the analysis, seeing, as the
insurance business is a good example of a service intensive sector. The relationship
between the EB, employee response, and SB is particularly relevant for companies
differentiation from competition, as most of the customers cannot distinguish
4 Empirical study 65
5 Research findings
All loadings except one (i.e., high job security) exceed 0.50 across "company
today" and "company of choice", and every indicator z-value was found to be
significant (see Table 1). Cronbachs a scores (Bagozzi and Yi, 1988) and average
variance extracted for all measurement scales indicate sufficient reliability. In
terms of convergent validity, the proportions of variance in the items explained by
the other factors are below the 0.50 criterion given by Fornell and Larcker (1981),
except social value and diversity value for "company today", both of which are at
0.524. In addition, the co-variance between the factors is lower than the variance in
the items explained by each of these, satisfying the discriminate validity criterion
(Fornell and Larcker, 1981). However, the variance of development value shared
with the factor social value for "company today" (0.438) is only slightly lower
than the variance in the items explained by the factor development value (0.446).
5 Research findings 67
Good reputation of the company amongst friends .70 (26.29) .66 (26.98)
Fit indices
c2 1839.57 1608.56
df 237 237
Notes:
z-Values shown in parentheses.
Table 2 shows the final results from the structural models. In the "company
today" model, c 2 is 2,082.905 with df = 276 and c 2 /d f ratio of 7.547, a CFI
of 0.930, and a RMSEA of 0.055 - which indicates that the model fits the data
well. The second structural model was assessed for "company of choice", which
indicates a better model fit concerning the Chi-square and comparative fit index:
c 2 = 1,703.45, df = 330; c 2 /d f = 6.654; CFI = 0.936; and RMSEA = 0.06. Not
all paths were supported in the model on "company today" since the z-values
associations with the structural path were not statistically significant (p < 0.10) or
the signs were incorrect. Similarly, two hypothesised paths had to be excluded
from the model of "company of choice", as the structural relationships were found
not to be significant.
The empirical results confirm that the perceived EB affects employees satisfaction
and identification with the company, as shown in Table 2. Testing the path from
economic value to employees job satisfaction at "company today" supports H1a
(b = 0.15, p < 0.01); however, the effect of economic value on identification with
the company, as hypothesised by H1b, cannot be supported, as the structural path
is not significant. Employees development value had a significantly positive
effect (b = 0.33, p < 0.01) on employees job satisfaction working with the
"company today" (H2a), whereas there was no significantly positive effect on
the identification with the company (H2b). In terms of perceived social value,
5.1 Current employees 69
Company today
H3a Social Value > Satisfaction + .20 *** Yes RMSEA .055
H4a Diversity Value > Satisfaction + .14 *** Yes TLI .917
Fit indices
Company of choice
H3c Social Value > Identification + .42 *** Yes RMSEA .060
H4c Diversity Value > Identification + .16 *** Yes TLI .925
H5c Reputation Value > Identification + .37 *** Yes CFI .936
Notes:
* p < 0.10; ** p < 0.05; *** p < 0.01.
H3a and H3b can be clearly supported. Social value is found to significantly
impact both employees job satisfaction (b = 0.20, p < 0.01) and identification
with "company today" (b = 0.19, p < 0.01). The empirical results support the
hypothesised significantly positive effect of "diversity value" on job satisfaction
H4a (b = 0.14, p < 0.01). On the other hand, the estimated effect on identification
with the company, as hypothesised by H4b, is not significant. Finally, in regard
to the effect of reputation value on employees job satisfaction and identification
with "company today", Table 2 indicates that the assumed effect on job satisfaction
cannot be supported. Thus, H5a is rejected. However, the path from reputation
value to identification with company today is significantly positive (b = 0.49, p <
0.01), supporting H5b.
It is obvious that, to support the creation of the service brand among current
employees, a mixture of measures that aim at current employees satisfaction as
well as measures that aim at current employees identification needs to be consid-
ered. For example, social value was proven to support both favourable employee
outcomes. Therefore, it is reasonable for companies to enhance this dimension.
We also suggest that the dimensions that have the strongest influence on identifica-
tion with the company and satisfaction should be taken into account when creating
the service brand. Concerning satisfaction, the dimension development value is
found to have the highest influence. In order to promote identification with the
company, reputation value in particular needs to be delivered to current employees.
Clearly, it is not an easy task to find the right proportion, as some dimensions
clearly support one outcome but not the other (i.e., development value).
employees expect as development value and economic value is not a positive driver
for employees identification with an employer of their choice. The expected
social value has the strongest positive effect on identification (b = 0.42, p < 0.01).
Therefore, H3c is supported. As proposed by H4c, what employees expect as
diversity value influences their identification with an employer (b = 0.16, p <
0.01). Thus, when considering potential employees that are important for the
long-term success of the service brand, the provision of development value and
reputation value in particular needs to be leveraged. Similarly, reputation value is
a strong positive driver for employees identification with "company of choice" (b
= 0.37, p < 0.01), as assumed by H5c. As these dimensions have also been proven
to provoke favourable employee outcomes among current employees, a company
should focus on them when creating the service brand through the EB.
6 Limitations/future research/discussion
Despite these limitations, our study provides valuable insights to enhance the
understanding of SB. A service brand is created in the triangle between company,
employee, and customer. We state that a strong EB is an efficient instrument for
fostering employee outcomes related to SB. It enhances employee outcomes in
such a way that they, in turn, influence customer experience positively and thereby
indirectly influence SB as follows:
Next, our investigation related the EB to the service profit chain. By measuring
employee satisfaction and employee identification with a company, both the
conventional and the complementary service profit chain (Homburg et al., 2009)
can be influenced by the EB. Therefore, our paper suggests that EB enhances
the well-researched service profit chain and indirectly pays off by leveraging a
companys profits (Anderson and Mittal, 2000; Heskett et al., 2003; Heskett et al.,
1994). As a result, companies may gain a competitive advantage by strengthening
the EB, which enables them to manage their employees skills and favourable
attitudes. This is strategically relevant to employee-customer interaction (Brodie
et al., 2009; Homburg et al., 2009; Vargo and Lusch, 2004; Vella et al., 2009).
Overall, the empirical results support that a strong EB enables companies to
influence current and prospective employees. The relation between the EB and
favourable employee outcomes is described empirically. Theoretically, we argue
that these employee outcomes, in turn, provoke favourable customer outcomes
that lead to the creation of the service brand. Thus, we propose the EB as a support
for SB activities.
As shown, the EB supports favourable employee attitudes that enhance the cus-
tomers experiences during interaction (Bernhardt et al., 2000; Bettencourt and
Brown, 1997; Hartline and Ferrell, 1996; Homburg and Stock, 2004; Homburg
et al., 2009; Pugh, 2001) and therefore hold a key role in the service setting.
Customers experiences during the service delivery process shape their image of
the companys brand (Berry et al., 2002; Brodie et al., 2009; Vargo and Lusch,
2004).
The research findings support the assumption that a companys ability to create
and to deliver an attractive image of the employer to the potential labour mar-
ket increases the likelihood of attracting high-potential applicants (Collins and
Stevens, 2002; Slaughter et al., 2004), which is of great significance to the com-
panys future. Therefore, it must be imperative for companies to foster those value
dimensions that attract potential employees: Social value, which encompasses the
social image of a company, including a respectful environment; friendly relation-
ships among co-workers; and a "people-first" attitude, which attracts employees
even more than the good reputation of products or company. Hence, companies
that support a strong culture of corporate social responsibility have an advantage
regarding competition for talented and identified employees. Apart from social
value, reputational aspects also (Cable and Graham, 2000; Edwards, 2010) in-
fluence potential employees identification with a company. This means that a
company needs to invest in a good reputation or well-known products. Hence,
service providers are challenged to build a good reputation in order to attract
identified high-potentials that may create the future service brand.
76 III S ERVICE B RANDING
8 Conclusion
Bibliography
Aaker, D. A., 1991. Managing brand equity: Capitalizing on the value of a brand
name. The Free Press, New York, NY.
Ambler, T., 2000. Marketing and the bottom line. Financial Times/Prentice Hall,
London, UK.
Ambler, T. and S. Barrow, 1996. The employer brand. Journal of Brand Manage-
ment, 4(3):185-206.
Avery, D. R., P. F. McKay, and D. C. Wilson, 2007. Engaging the aging workforce:
The relationship between perceived age similarity, satisfaction with coworkers,
and employee engagement. Journal of Applied Psychology, 92(6):1542-1556.
Berry, L. L., L. P. Carbone, and S. H. Haeckel, 2002. Managing the total customer
experience. MIT Sloan Management Review, 43(3):1-17.
Bitner, M. J., 1995. Building service relationships: Its all about promises. Journal
of the Academy of Marketing Science, 23(4):246-251.
Burmann, C., K. Schaefer, and P. Maloney, 2008. Industry image: Its impact on
the brand image of potential employees. Brand Management, 15(3):157-176.
Carmeli, A., 2005. Perceived external prestige, affective commitment, and citizen-
ship behaviors. Organization Studies, 26(3):443-464.
Davies, G., R. Chun, R. da Silva, and S. Roper, 2004. A corporate character scale
to assess employee and customer views of organization reputation. Corporate
Reputation Review, 7(2):125-146.
Grnroos, C., 2006. Adopting a service logic for marketing. Marketing Theory,
6(3):317-333.
Harrison, D. A., D. A. Newman, and P. L. Roth, 2006. How important are job
attitudes? Meta-analytic comparisons of integrative behavioral outcomes and
time sequences. Academy of Management Journal, 49(2):305-325.
82 III S ERVICE B RANDING
Hatch, M. J. and M. Schultz, 2001. Are the strategic stars aligned for your
corporate brand. Harvard Business Review, 79(2):128-134.
Heskett, J. L., W. E. Sasser, and L. A. Schlesinger, 1997. The service profit chain:
How leading companies link profit and growth to loyalty, satisfaction, and
value. The Free Press, New York, NY.
Heskett, J. L., W. E. Sasser, and L. A. Schlesinger, 2003. The value profit chain:
Treat employees like customers and customers like employees. The Free Press,
New York, NY.
Hoek, J., J. Dunnett, M. Wright, and P. Gendall, 2000. Descriptive and evalua-
tive attributes: What relevance to marketers? Journal of Product & Brand
Management, 9(6):415-435.
Homburg, C. and R. M. Stock, 2004. The link between salespeoples job satis-
faction and customer satisfaction in a business-to-business context: A dyadic
analysis. Journal of the Academy of Marketing Science, 32(2):144-158.
Homburg, C., Wieseke, J. and W. D. Hoyer, 2009. Social identity and the service-
profit chain. Journal of Marketing, 73(2):38-54.
B IBLIOGRAPHY 83
Hsieh, M. H., S. L. Pand, and R. Setiono, 2004. Product-, corporate-, and country-
image dimensions and purchase behavior: A multicountry analysis. Journal
of the Academy of Marketing Science, 42(3):251-270.
Judge, T. A., J. E. Bono, and E. A. Locke, 2000. Personality and job satisfaction:
The mediating role of job characteristics. Journal of Applied Psychology,
85(2):237-249.
Jurgensen, C. E., 1978. Job preferences (What makes a job good or bad?). Journal
of Applied Psychology, 63(3):267-276.
Kim, C. K., D. Han, and S. B. Park, 2001. The effect of brand personality
and brand identification on brand loyalty: Applying the theory of social
identification. Japanese Psychological Research, 43(4):195-206.
Kimpakorn, N. and G. Tocquer, 2010. Service brand equity and employee brand
commitment. Journal of Services Marketing, 24(5):378-388.
84 III S ERVICE B RANDING
King, C. and D. Grace, 2005. Exploring the role of employees in the delivery of the
brand: A case study approach. Qualitative Market Research: An International
Journal, 8(3):277-295.
Lievens, F., G. Van Hoye, and F. Anseel, 2007. Organizational identity and
employer image: Towards a unifying framework*. British Journal of Manage-
ment, 18(s1):45-59.
Mael, F. A. and B. E. Ashforth, 1992. Alumni and their alma mater: A partial
test of the reformulated model of organizational identification. Journal of
Organizational Behavior, 13(2):103-123.
Malka, A. and J. Chatman, 2003. Intrinsic and extrinsic work orientations as mod-
erators of the effect of annual income on subjective well-being: A longitudinal
study. Personality and Social Psychology Bulletin, 29(6):737-746.
Merz, M. A., Y. He, and S. L. Vargo, 2009. The evolving brand logic: A service-
dominant logic perspective. Journal of the Academy of Marketing Science,
37(3):328-345.
Michaels, E., H. Handfield-Jones, and B. Axelrod, 2001. The war for talent.
Harvard Business School Press, Boston, MA.
Mosley, R. W., 2007. Customer experience, organisational culture and the em-
ployer brand. Journal of Brand Management, 15(2):123-134.
Payne, A., K. Storbacka, P. Frow, and S. Knox, 2009. Cocreating brands: Diagnos-
ing and designing the relationship experience. Journal of Business Research,
62(3):379-389.
Pugh, S. D., 2001. Service with a smile: Emotional contagion in the service
encounter. Academy of Management Journal, 44(5):1018-1027.
Ritson, M., 2002. Marketing and HE collaborate to harness employer brand power.
Marketing, 24(10):24.
Smidts, A., A. T. H. Pruyn, and C. B. M. van Riel, 2001. The impact of employee
communication and perceived external prestige on organizational identifica-
tion. Academy of Management Journal, 44(5):1051-1062.
B IBLIOGRAPHY 87
Weathington, B. L., 2008. Income level and the value of non-wage employee
benefits. Employee Responsibilities and Rights Journal, 20(4):291-300.
88 III S ERVICE B RANDING
Wiertz, C., K. de Ruyter, C. Keen, and S. Streukens, 2004. Cooperating for service
excellence in multichannel service systems: An empirical assessment. Journal
of Business Research, 57(4):424-436.
Wilden, R., S. Gudergan, and I. Lings, 2010. Employer branding: Strategic impli-
cations for staff recruitment. Journal of Marketing Management, 26(1/2):56-
73.
89
Abstract
The increasing attractiveness of emerging markets (EMs), along with the progress-
ing internationalization raises the question whether setting up a unified marketing
strategy for EMs and high income countries (HICs) is appropriate. In this paper,
the requirements of an international segmentation including EMs are elaborated.
The established requirements are numerically tested by applying multilevel finite
mixture modeling on global consumer multichannel (search and purchase) behav-
ior, thereby providing an initial large-scale study that compares consumers from
EMs and HICs. In this course, existing finite mixture models are enhanced to
reflect institutional effects, which are deemed tremendous in EMs, by combining
information on consumer channel usage patterns with consumer and contextual
covariates. Findings broadly support the suggested requirements, contributing to
international segmentation and EMs literature. Practitioners can directly use the
insights of this study for establishing an international marketing strategy, whereas
the key finding proposes that a case-by-case analysis for EMs is necessary and
that solely considering EMs as such derives wrong conclusions about international
segmentation.5
5 T. Schlager and P. Maas. Previous versions of this paper were accepted for presentation and
the American Marketing Association Summer Marketing Educators Conference 2012, and the 41st
Annual Conference of the European Marketing Academy 2012. At the latter conference, it has been
nominated for the award "Best paper based on a doctoral dissertation". Further, it has been handed in
at the "Journal of International Marketing" where it has yet advanced to the second review round.
90 IV E MERGING M ARKETS
1 Introduction
The fast rise of emerging economies has driven a shift whereby the
centers of economic growth are distributed across developed
and developing economies - its a truly multipolar world
- Justin Yifu Lin, the World Banks chief economist and senior
vice president for development economics (2011)
In line with both examples, scholars see reasonable evidence to state that
consumers from EMs substantially differ from those in well-investigated HICs
(Burgess and Steenkamp, 2006; Sheth, 2011). Accordingly, taking in the MNCs
perspective, Sheth (2011, p. 176) explicitly questions, "should a company extend
or adjust its marketing mix to suit the local [emerging] markets"? To answer this,
MNCs must have a thorough understanding of EMs consumers preferences and
especially, of the differences to HICs consumers since ignoring their peculiarities
can end up in long-enduring difficulties up to complete failures.
92 IV E MERGING M ARKETS
The purpose of this article is to fill these gaps by shedding light on how
EMs and HICs consumers differ; specifically, an approach to international
segmentation that includes EMs is established. Primarily, we argue for splitting
countries in smaller pieces according to geographical and class-specific patterns,
accompanied by using distinct variables to detect institutional influences. The tone
of this article is neither to "Easternize", which is to solely focus on EMs, nor to
"Westernize", but rather to provide a balanced view on international segmentation.
2 Prior literature on EMs and segmentation 93
For realizing these goals, we structure the article as follows. First of all, we
briefly delineate conceptual and empirical approaches to international segmenta-
tion and EMs. Next, we discuss the peculiarities of EMs and elaborate the specific
requirements and foci of an international segmentation that includes EMs. Thirdly,
we extend finite mixture modeling for our purpose and present a numerical exam-
ple on consumer channel behavior. Finally, we reflect on our study by discussing
findings, theoretical and managerial implications, as well as avenues for future
research.
For instance, Peterson and Malhotra (2000) cluster countries by measuring con-
sumers quality of life perceptions. They indicate their usefulness for international
segmentation and provide the largest segmentation study with 165 countries world-
wide so far. An important implication is that rather basic constructs tend to be
applicable for segmenting HICs and EMs. In line with this, Agarwal, Malhotra,
and Bolton (2010) found cross-culture-segments in their comparison of consumers
service quality perceptions in India and the US regarding their perceptions of
service quality. These insights confirm the usefulness of culture as segmentation
variable for EMs. Gaston-Breton and Martn (2011) provide an auspicious ap-
proach that first segments countries based on economic indicators, such as the
market size and potential. In a second step, consumers are clustered based on
Ingleharts values, accentuating the notion of basic values to be useful in interna-
tional segmentation. Cleveland, Papadopoulos, and Laroche (2011) find ethnic
identity and cosmopolitanism to be stable across cultures and countries. Moreover,
they provide one of the few studies that include EMs (India and Mexico), although
EM-specific implications are not drawn. All of the delineated studies include
EMs, nevertheless, none has directed the focus of attention on the requirements of
an approach that includes EMs.
ity; in general, the context differs across regions (Sheth, 2011). The prevailing
dissimilarity among consumers poses a severe threat to a central requirement of
segmentation, namely "segment homogeneity" (Wind, 1978). On consumer side,
geographical mobility is decisive in determining homogeneous segments. In an
early approach, Andreasen (1966) claims that "other consumptions patterns [. . . ]
are appropriate" in different geographical regions (p. 346). As economic and
diffusion literature states, mobility is needed for developing similar behaviors and
consumption habits, whereas dissimilarities arise as a result of immobility (e.g.,
Agarwal, 2003). Diverging consumer preferences and behaviors are more likely
to develop across the regions of EMs since the lower class, also referred to as
"bottom-of-the-pyramid", is limited in terms of their mobility (Cui and Liu, 2001)
and their possibilities for exchange. For instance, the lower class has little access
to transportation, electricity, or internet (Sheth, 2011). While they are forced to
buy in local stores, members of the upper class may be free to make a choice.
Missing exchange among consumers and immobility in particular unfold their
influence in large EMs, such as China and India (e.g., Enderwick, 2009), due
to their sheer size. Taking both as examples, more than 1 billion people live in
each country, thereby outstripping Europe with only about 700 million people
by far and paired with the limited ability to change location, geography-specific
consumption patterns tend to develop. Further, there is a large gap between the
upper class, which is assumed to converge to western consumer behavior, while
the lower class is particularly limited in their behavior, which underscores the
within-country heterogeneity (Alden et al., 2006; Burgess and Steenkamp, 2006).
Finally, an issue closely related to geography is language. For instance, China
has seven official languages; however, more than 80 languages are used on a
general basis. The number of different accents in conversational Chinese is even
higher, which introduces a further source of heterogeneity (e.g., Craig and Dou-
glas, 2006). Accordingly, EMs deviate from HICs in their level of within-country
heterogeneity.
98 IV E MERGING M ARKETS
Similar to the micro-level analysis, which opens avenues for finding global con-
sumer segments, variables and covariates for an international segmentation should
be thoroughly chosen in order to detect patterns underlying the segments. In
particular, we advocate construct equivalence, the focus on stable variables, and
the inclusion of contextual covariates.
Construct Equivalence
Myers et al. (2000) as well as Steenkamp and Ter Hofstede (2002) emphasize
the importance of construct equivalence for international studies, which refers
to whether the variables used for segmentation differ in their meanings across
countries (Steenkamp and Ter Hofstede, 2002). The rationale behind a focus on
construct equivalence in international segmentation is that "segmentation criteria
for one market may not work in another" (Cunningham and Crissy, 1972, p. 100).
While this is crucial for international studies, an even higher focus needs to be laid
on construct equivalence when EMs are included in the sample since there is rea-
sonable evidence that EMs significantly differ in their decision-making processes
(e.g., Burgess and Steenkamp, 2006). To date, only few constructs have been
confirmed to be useful for cross-border research (Deshpand, 1999). In reflecting
this idea, Sheth (2011) points out that prior to deriving constructs, scholars first
need to investigate actual consumer behavior, which is then suggested to be a solid
segmentation basis. Sociodemographic variables, income and the social class
(Burgess and Steenkamp, 2006; Cunningham and Crissy, 1972; Kamakura and
Wedel, 1995; Sheth, 2011) cannot only be used for defining micro-level analysis,
but may also serve as covariates, depending on the segmentations objectives. In
4.3 Segmentation variables and covariates 101
Brazil, for instance, Cunningham et al. (1974) found distinct shopping behaviors,
such as the product selection, the relevant services, the information sources, and
the transportation system used for arriving at the store, depending on the class a
consumer belongs to. Consistently, Steenkamp and De Jong (2010) recognized
that demographic variables are important indicators of Western behavior in EMs.
However, as shown by recent research (Strizhakova et al., 2012; Strizhakova et
al., 2011), patterns might need to be adapted. In studies on Russia and Brazil,
Strizhakova et al. (2012) found four different segments within the young adult
cohort; two of them, the "glocally-engaged" and the "globally-engaged", are
particularly prone to global firms and brands. Thinking in cohorts rather than in
age might therefore be more germane in EMs concerning the affinity to MNCs
products. Reviewing previous segmentation studies, several further types of vari-
ables have been developed for cross-border segmentation, among them attitudes
and perceptions (e.g., Agarwal et al., 2010; Lemmens et al., 2007; Peterson and
Malhotra, 2000; Ter Hofstede et al., 1999; Yavas et al., 1992), and values (e.g.,
Gaston-Breton and Martn, 2011), which have thus been confirmed regarding
construct equivalence.
The rapidly changing context in EMs is not solely relevant in the selection of
respondents; a considerable challenge is presented by finding a stable segmenta-
tion basis, or by controlling for dynamics. To a certain degree a trade-off between
considering rapid changes and selecting solid segmentation variables is inevitable
when including EMs. Yet, research has identified variables that are fairly stable
and thus, can be drawn on for an international segmentation that includes EMs.
An individuals value system not only shows high construct equivalence, it also
tends to remain rather stable over time (Bilsky and Schwartz, 1994) as it reflects
an individuals fundamental beliefs. Moreover, values contain considerable ex-
planatory power for consumption and purchase behavior (Burgess, 1992). The
most prominent system to measure values stems from Schwartz (1992), which
has been extensively applied and tested by international research and for EMs
102 IV E MERGING M ARKETS
a characteristic that coins the context and has been identified as influential by a
variety of empirical and conceptual papers is culture (e.g., Douglas and Craig,
1997, 2006; Hsieh et al., 2004). Several frameworks for classifying countries
according to culture are suggested (Hofstede, 1983, 1991; Schwartz, 1992), which
can be drawn on for EMs. To wrap up, the context needs to be reflected when
determining international segments.
Segmentation studies have the inherent aim to aggregate consumers and to simplify
the reality in order to find feasible managerial approaches. However, in line
with Sheth (2011) who questions the possibility of segmentation approaches,
we remind researchers not to oversimplify the real situation as the approaches
for EMs may not arrive at a similar simple level as in HICs. Referring to the
fragmented supply side, the highly heterogeneous consumer market, and the
contextual embeddedness, an abstraction of the reality may fail to provide a
reliable picture of a markets landscape. Consistently, Burgess and Steenkamp
(2006) name the presence of confounding institutional effects in EMs, which may
result in wrong inferences about the segment membership. Complex interactions
could disguise the real reason for consumer behavior; therefore, unmasking these
relationships is particularly relevant for the success of a segmentation that includes
EMs. On an individual level, Burgess and Steenkamp (2006) exemplify this by
means of shopper profiles: Data collected by supermarket scanners tends to be
highly heterogeneous in EMs, even from the same location. In case the consumers
profiles are connected to systemic differences, one may accumulate shopper and
purchase data to control for patterns that influence the purchase. On a contextual
level, the slow technology adoption in EMs serves as example. Whereas one may
attribute the slow technology adoption to the cultural or technological context,
the reason might be of regulative nature (Calantone et al., 2006). Attaching
the development to cultural aspects only, would fail to reveal the underlying
mechanism. Accordingly, MNCs need to determine the sometimes-unobvious
104 IV E MERGING M ARKETS
when using primary data from EMs. To enhance translation equivalence, common
approaches are presented by backtranslation or parallel translation (Douglas and
Craig, 2006).
Conceptual level
Level of analysis Macro-level analysis > Micro-level analysis
a) Regional / statewise analysis
b) Area-level analysis (rural vs. metropolitan)
c) Class-level analysis (upper class vs. bottom-of-the-pyramid)
Selection of relevant respondents Individuals > Relevant respondents (to unmask decision-making processes)
a) Cautious reliance on individuals
b) Consideration of peer-group leaders (word-of-mouth)
c) Focus on group-effects (consumer interactions)
Segmentation variables and covariates Construct equivalence > Careful variable and covariate selection
a) Emphasis on construct equivalence (lack of variables / covariates)
b) Evaluation of variables / covariates temporal stability
c) Inclusion of contextual covariates
Extent of the approach Focused approach > Holistic approach
a) Adoption of individual-level covariates
b) Adoption of contextual covariates
c) Consideration of interactions among variables / covariates
Methodological level
Primary data Measurement equivalence > Valid and reliable primary data
a) Emphasis on measurement equivalence
IV E MERGING M ARKETS
Translation equivalence
Calibration equivalence
Score equivalence
b) Usage of simplified research settings / constructs
Context level data Standardized global data > Comparable Secondary Micro-level Data
a) Inclusion of comparable micro-level data
b) Evaluation of reliability of micro-level data
c) Data adjustment (calibration equivalence)
segment membership with respect to the binary data structure at hands. For de-
riving the basic model, we borrowed from Bijmolt et al. (2004). In accordance,
let
The consumer segment membership represents the latent trait, Xi j for each con-
sumer of a higher-level group using a specific combination of channels. Equally,
Z j describes the higher-level segment membership. Equation (1) denotes the
conditional probability of using a specific combination of channels for searching
and buying Yi j for a consumer, depending on the membership of higher-level
group to a higher-level segment (Vermunt, 2001). N j represents the sample size of
a higher-level group. The respective equation of the multilevel mixture model is
specified as follows:
" " ##
T Ni S K
P(Y j ) = P(Z j = t) P(Xi j = s|Z j = t) P(Yi jk |Xi j = s) (1)
t=1 i=1 s=1 k=1
5.2 Basic model 111
L
exp g0s0 t + gls0 Vli j
l=1
P(Xi j = s0 |Z j = t) = S
L
(2)
exp g0st + glsVli j
s=1 l=1
This delineation is consistent with the basic multilevel finite mixture model
as proposed for international segmentation by Bijmolt et al. (2004). For a more
detailed model delineation please refer to their article. To reflect the scholarly
emphasis on the impact of the context (e.g., Burgess and Steenkamp, 2006;
Douglas and Craig, 2011), we extended the model in a way that allows for
including higher-level covariates. Again, the covariates were assumed to directly
influence the higher-level segments. In this respect, let W m j be the higher-level
covariate (with 1 m M for the higher-level covariates). We delineate equation
(3):
112 IV E MERGING M ARKETS
L
exp g0s0 t 0 Wm j + gls0 Vli j
l=1
P(Xi j = s0 |Vli j , Z j = t 0 |Wm j ) = S
L
(3)
exp g W
0st m j + g
ls li j
V
s=1 l=1
6 Numerical application
The study was conducted at the end of 2010. For collecting data, we employed
online surveys, which were broadly discussed and double-checked with bilingual
country representatives to avoid ambiguous descriptions among countries and
languages in line with our suggestions. For reaching consumers, we worked
together with local research contacts to guarantee reliable information (Hoskisson
et al., 2000). After initial questions, the channels were described in detail to
ensure a common understanding among respondents (Burgess and Steenkamp,
2006; Steenkamp and Ter Hofstede, 2002). At the end of the questionnaire,
respondents were asked questions regarding demographic patterns (age, gender,
living area), which were used as covariates. As psychographic measure we used
the three of the NEO-PIR personality traits, namely conscientiousness, openness,
and agreeableness, validated by Rammstedt and John (2007) which constitutes a
temporarily stable measure. We chose those three as they are assumed to relate to
consumer multichannel behavior. For instance, openness might be an indicator
for using personal channels since consumers are likely to find more pleasure
in interacting with others. Similarly, conscientious consumers are proposed
to gather information through multiple channels rather than relying on solely
one channel. Additionally, we prechecked the distribution of other covariates
regarding comparability. This provided us with 18,239 useful questionnaires from
five continents (Asia, Australia, Europe, North-America and South-America),
including six EMs: Brazil, China, India, Mexico, Poland, and the Czech Republic,
6.1 Data collection and analysis 113
which represent EMs from all continents but Africa. Though no EM from Africa
was included, we argue that the sample is appropriate as it includes multiple EMs
that have a considerably different context.
Before analyzing segments, data was weighted within (to correct for deviances
of the respective countrys demography by case weights) and between countries
(to obtain inferences about real segment sizes) (Schaffer and Riordan, 2003;
Steenkamp and Ter Hofstede, 2002) according to age, gender, income and level
of urbanization. The descriptive results of the consumers choice regarding the
search and purchase channel are displayed in Appendix A.
114 IV E MERGING M ARKETS
6.2 Results
Segment 1 UK
2 Singapore, Malaysia, India low
3 Spain, Korea, France, Brazil (high and low),
Mexico (high, medium, low)
4 Japan (high, medium, low), Germany (high, medium)
5 China (high , medium, low)
6 US (high), Poland (high, medium, low),
CZ (high, medium, low), Germany (low)
7 Austria, Belgium, India (high, medium), US (middle, low)
8 Australia, Sweden, NL
At first glance, the large number of consumer segments (i.e., 14) does not
appear practicable. However, we argue differently for three reasons. First, in
our study consumers were able to specify behavior from eight search and four
buying channels. For example, if only the combination of one search channel
was chosen, 32 possible combinations arose. To realistically describe segments,
consumers were able to check up to three channels for the search phase (i.e.,
searching via a providers website and through a personal contact, while buying
via an independent web provider) which theoretically leaves room for thousands
of unique combinations. Second, within several higher-level segments, specific
consumer segments nearly do not exist, which fairly reduces the number of
segments to be consider for a specific market (Bijmolt et al., 2004). Finally,
five consumer segments (existing in nearly all regional segments) account for 62
percent of the consumers. We conclude that the number of consumer segments
does not affect the practicability of our results.
We used three values to assess how well our model predicts class membership.
First, as suggested by previous studies, we applied an entropy measure which
116 IV E MERGING M ARKETS
consumers. Table 6 reports the probability for any consumer segment to appear in
any higher-level segment.
118
Channel usage probabilities x Channel usage probabilities
Channels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7 8 Overall
Size .06 .09 .08 .13 .12 .14 .08 .11 .13 .05 .03 .08 .21 .13 .08 .22 .16 .08
Onl. provider .94 .35 .11 1 .10 .81 .64 .76 .40 1 .11 .32 .34 .61 .31 .52 .51 .61 .43 .44 .64 .49
Onl. independent .12 .99 .09 .40 .06 .27 .86 .17 .87 .56 .39 .41 .67 .21 .24 .51 .29 .41 .32 .40 .37
Search
Forum .23 .07 .07 .15 .91 .14 .08 .13 .09 .08 .86 .41 .11 .17 .12 .08 .19 .16 .10 .10 .13
Offl. provider .61 .36 .98 .33 .98 .98 .12 .65 1 .18 .45 .69 .32 .69 .71 .51 .64 .55 .57 .52 .58
Offl. independent .55 .41 .21 .10 .98 .95 .08 .19 .11 .94 .11 .95 .27 .47 .36 .31 .39 .41 .49 .26 .38
Peers .46 .45 .89 .29 .41 .20 .44 .25 .41 1 1 .58 .47 .45 .36 .57 .45 .43 .49 .50 .47 .39 .46
Ads .07 .04 .12 .17 .10 .06 .16 .47 .11 .06 .31 .14 .14 .14 .14 .09 .11 .13 .16 .13
Journals .14 .09 .10 .10 .10 .22 .15 .16 .12 .10 .10 .11 .05 .12 .11 .12 .11
Onl. provider .23 .09 .72 .05 .41 .10 1 .13 .30 .07 .15 .20 .23 .15 .15 .29 .18
Purchase
Onl. independent .24 .06 .53 .10 .06 .15 .29 .04 .05 .19 .06 .11 .07 .12 .11
Offl. provider .50 .41 .87 .19 .64 1 .91 .04 .74 .99 .21 .59 .27 .65 .63 .45 .52 .51 .50 .46 .51
IV E MERGING M ARKETS
Offl. independent .25 .25 .12 .33 .03 .07 .71 .13 .98 .14 .24 .17 .16 .19 .24 .28 .14 .20
Notes:
The largest probability and relative size is underlined (for searching and buying).
The second largest value is boldface (for searching and buying).
Probabilities below .03 not shown.
Onl. = online; Offl. = offline; Ads = advertisements; independent = not directly related to the company bought from.
Consumer segments
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Notes:
Across the country segments, the largest probability and relative size is underlined, the second
largest value is bold face.
Consumer segments below .03 in a higher-level segment not shown.
Effects of Covariates
(a)
(b)
Notes:
Only the segments with more than 6 % of the actual consumers shown.
that high-income classes of EMs are consistently more similar to HICs consumers
cannot be fully supported: The upper class of India, for instance, appears in one
segment with several HICs. Similarly, the lower class of Germany is classified
in one segment with several EMs; however, also the upper class of the US is
registered into this higher level segment. Although this is counterintuitive, it
may be explained when considering the development of usage patterns and of
the industry (Burgess and Steenkamp, 2006). The lower class consumers in EMs
might use the personal channel for buying financial services as they have not
accommodated to using Internet for such purchases yet. This, however, is similar
to the upper class in HICs, where personal contact is often used due to the better
service. Thus, as consumer preferences regarding how to search and buy financial
services are still developing in EMs, which suggests including the development of
the specific industry as another variable. Although we specified the micro-level
analysis based on classes, we solely asked consumers that already have Internet
access, which attaches higher homogeneity to consumers worldwide. While this
7 Discussion and implications 123
Additionally, EMs did not consistently show up in one segment which indicates
that there is considerable heterogeneity between different EMs. This strongly
advocates against considering EMs as homogeneously different from HICs, but
rather indicates that scholars need to examine them on a case-by-case level. The
notion of embeddedness and the holistic model was reflected by considering
contextual factors and the high number of covariates. Their usefulness was clearly
supported since all contextual variables consistently improved the segmentation.
Although at first sight not all segments relate to each other, considering covariates
explains the higher segments as illustrated by Figure 9.
It is noteworthy that the numerical example provides a first step into EMs
research, which has been strongly advocated by multiple scholars (e.g., Burgess
and Steenkamp, 2006). In line with Sheth (2011), we relied on consumer behavior
as to date other constructs miss a validation for EMs. Thereby, we enhance
the understanding about international channel segments and their channel usage
patterns. In recognizing that different segments exist in different countries, and
that country segments are not as intuitive as expected, marketing practitioners may
rethink current segmentation approaches. Equally, the applied covariates have yet
been specified as valid for research on an international level. For instance, we
used psychographic elements that have been suggested to be globally relevant by
including the five personality traits into the segmentation (Burgess and Steenkamp,
2006). As a result, we were able to better fit the segmentation.
(e.g., Bijmolt et al., 2004; Ter Hofstede et al., 1999), which we further extended
by considering contextual information. In our example, the inclusion of covari-
ates that are allowed to directly affect higher-level segments, ends in deriving a
significantly better model which underlines their usefulness.
MNCs have yet acknowledged the potential of EMs. Since decades, they strive to
expand their businesses beyond the boundaries of HICs to elude fierce competi-
tion in HICs and to exploit the extensive growth rates in EMs (Prahalad, 2004).
Specifically, MNCs should determine global segments based on class-specific
as well as geographical patterns, whereas using a "country-as-a-segment" strat-
egy is inferior. However, we need to emphasize that treating all EMs equally
should be reconsidered. To acknowledge that EMs are highly diverse is essential
to establish a successful international marketing or internationalization strategy.
One additional key issue becomes obvious when considering the results of our
segmentation: Grouping countries according to continents is not effective. This
has direct implications on the MNCs organization structure, which should reflect
the markets heterogeneity. Predominantly, internal structures are grouped based
on geographic proximity (i.e., PUMAs organization is split into North America,
Americas, Central Europe, Eastern Europe Middle East Africa, and Asia- Pacific),
which is, according to our results, inferior. Further, a key question for MNCs that
7.2 Future research 125
The findings and ideas of this article can guide future studies. In the beginning,
we split countries into three regions to detect within-country heterogeneity (e.g.,
Burgess and Steenkamp, 2006; Cui and Liu, 2000), which leaves space for finer-
126 IV E MERGING M ARKETS
grained research. For instance, one might conduct research on a statewise level,
which could give further insights about the geographic segment distribution. Then,
we segmented based on consumers class membership, which displayed superior
results. Future research might combine both classifications to derive even more
specific results for unmasking the source of heterogeneity in EMs. Additionally,
our sample misses EMs from Africa. Previous research has shown interesting
findings for this continent (e.g., Steenkamp and Burgess, 2002), advocating the
inclusion of African countries. Our methodology revised current multilevel finite
mixture models by including contextual covariates. Particularly, we specified
them equivalently to consumer covariates (Gupta and Chintagunta, 1994). Our
approach showed to be useful in terms of model fit, but specific research dedicated
at further testing concurrent approaches is necessary. For example, an interesting
question is whether a two-step approach that first estimates on a consumer level,
and subsequently forms higher-level segments with covariates, is superior to our
approach.
A Descriptive statistics
Australia 654 .55 .40 .06 .53 .29 .41 .26 .07 .26 .08 .52 .16
Austria 612 .40 .33 .08 .63 .58 .44 .06 .11 .05 .07 .54 .35
Belgium 634 .51 .21 .08 .61 .51 .47 .09 .11 .11 .04 .54 .31
Brazil 1,406 .51 .18 .15 .70 .37 .44 .17 .12 .16 .04 .64 .16
China 1,372 .62 .29 .23 .64 .40 .40 .07 .03 .20 .06 .55 .19
Czech Republic 757 .42 .40 .15 .58 .41 .58 .06 .15 .16 .08 .50 .26
France 697 .52 .37 .09 .72 .21 .36 .11 .12 .14 .05 .73 .09
Germany 1,313 .48 .52 .12 .51 .41 .38 .06 .17 .15 .17 .45 .22
India 1,316 .27 .22 .14 .68 .46 .52 .29 .14 .08 .06 .56 .29
Japan 1,378 .50 .51 .05 .52 .28 .46 .18 .09 .20 .24 .43 .13
Korea 707 .47 .23 .15 .66 .32 .53 .20 .06 .19 .07 .59 .15
Malaysia 713 .38 .15 .21 .73 .38 .64 .16 .08 .07 .04 .69 .20
Mexico 672 .60 .12 .06 .80 .41 .46 .13 .12 .12 .02 .65 .20
Netherlands 597 .69 .25 .09 .44 .35 .44 .10 .11 .26 .11 .40 .22
Poland 664 .32 .42 .25 .63 .35 .53 .10 .13 .14 .06 .58 .22
Singapore 679 .32 .22 .16 .66 .53 .62 .11 .04 .06 .04 .59 .30
Spain 675 .58 .26 .15 .64 .44 .46 .11 .07 .19 .07 .53 .20
Sweden 685 .72 .49 .12 .54 .20 .38 .07 .13 .29 .15 .51 .05
UK 1,281 .60 .70 .09 .28 .28 .34 .20 .07 .33 .27 .26 .14
USA 1,427 .41 .38 .08 .51 .48 .47 .14 .08 .19 .10 .43 .27
Total 18,239 .49 .35 .13 .59 .38 .46 .14 .10 .17 .10 .52 .20
Notes:
EMs in bold, independent = not directly related to the company bought from.
Ads = advertisements.
127
128 IV E MERGING M ARKETS
China: We used the article of Cui and Lui (2000) that split China in seven regions. However,
we further simplified the classification and only accounted for three regions: North (formerly
North-West, North and North-East), East, and South (formerly South-West, South and Central)
China.
India: Similarly, India was split into three parts based on the classification of Husain, Dutta,
and Ghosh (2011). Hence, we classified India in South, North (formerly North, Central and
East), and West India. We acknowledge splitting both countries into three geographic regions
as reasonable, as the biggest difference in China and India prevail between northern and
southern parts (Douglas and Craig, 2011).
Brazil: Brazil was split in three parts, North (reaching to Gols in the south), south (reaching
up to Minas Gerais in the north) and the metropolitan area (Sao Paolo and Rio de Janeiro).
Note: To also investigate potential heterogeneity of HICs, we also split Germany, Japan, and USA into
regional segments.
Japan: Japan was split into the Northern, Southern part. Further, the district Kanto (the region
around Tokyo) was separated.
USA: The USA was split into West, Central, and East. Central was formed by Alabama,
Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Michigan, Minnesota, Missis-
sippi, Missouri, Nebraska, Ohio, Oklahoma, North and South Dakota, Tennessee, Texas and
Wisconsin. All states Western to this were classified as West, all Eastern to this as East.
Germany: Germany was divided into the Western part (former Federal Republic of Germany)
and the Eastern part (former German Democratic Republic).
Each country used for the geographical classification was split into three parts, the upper class,
the medium and the lower class.
Bibliography
Bolton, L. E., H. T. Keh, and J. W. Alba, 2010. How do price fairness perceptions
differ across culture? Journal of Marketing Research, 47(3):564-576.
Bozdogan, H., 1987. Model selection and Akaikes information criterion (AIC):
The general theory and its analytical extensions. Psychometrika, 52(3):345-
370.
130 IV E MERGING M ARKETS
Burgess, S. M. and J.-B. E. M. Steenkamp, 2011. Call for papers - Special issue on
marketing in emerging markets. (Available at http://www.elsevierscitech.com).
Census Bureau United States, 2011. 2010 Census data. (Available at census.gov).
Clark, T., 1990. International marketing and national character: A review and
proposal for an integrative theory. Journal of Marketing, 54(4):66-79.
Dekimpe, M. G., 2009. From the (new) editor. International Journal of Research
in Marketing, 26(4):259-260.
Enderwick, P., 2009. Large emerging markets (LEMs) and international strategy.
International Marketing Review, 26(1):7-16.
Griffith, D. A., 2012. Call for Papers: Internationalization during times of chang-
ing markets. (Available at http://www.marketingpower.com).
Hsieh, M. H., S. L. Pan, and R. Setiono, 2004. Product-, corporate-, and country-
image dimensions and purchase behavior: A multicountry analysis. Journal
of the Academy of Marketing Science, 32(3):251-270.
Husain, Z., M. Dutta, and S. Ghosh, 2011. Contraceptive use among illiterate
women in India: Does proximate illiteracy matter? In MPRA paper. Munich:
University library Munich, Germany.
Johnson, J. and G. J. Tellis, 2008. Drivers of success for market entry into China
and India. Journal of Marketing, 72(3):1-13.
Lee, C., and R. T. Green, 1991. Cross-cultural examination of the Fishbein behav-
ioral intentions model. Journal of International Business Studies, 22(2):289-
305.
Lin, J. Y., 2011. Emerging market growth poles are redefining global economic
structure. (Available at www.worldbank.org).
Prahalad, C. K., 2004. The fortune at the bottom of the pyramid: Eradicating
poverty through profits. Wharton School Publishing, Upper Saddle River, NJ.
Schwartz, S. H., 1992. Universals in the content and structure of values: The-
oretical advances and empirical tests in 20 countries. In M. Zanna, editor,
Advances in experimental social psychology, pages 1-65, Academic Press,
Orlando, FL.
Schwartz, S. H., 2004. Mapping and interpreting cultural differences around the
world. In H. Vinkin, J. Soeters, and P. Ester, editors, Comparing cultures: Di-
mensions of culture in a comparative perspective, pages 43-73, Brill Academic
Publishers, Leiden, Netherlands.
Sehgal, V., S. Lauster, E. Mayr, and G. Panneer. 2010. Roasted or fried - How to
succeed with emerging market consumers. Booz & Company. (Available at
www.booz.com).
Strizhakova, Y., R. A. Coulter, and L. L. Price, 2012. The young adult market
in emerging markets: Assessing their glocal cultural identity in a global
marketplace. International Journal of Research in Marketing. 29(1):43-54.
Talukdar, D., K. Sudhir, and A. Ainslie, 2002. Investigating new product diffusion
across products and countries. Marketing Science, 21(1):97-114.
Ter Hofstede, F., J.-B. E. M. Steenkamp, and M. Wedel, 1999. International market
segmentation based on consumer-product relations. Journal of Marketing
Research, 36(1):1-17.
Ter Hofstede, F., M. Wedel, and J.-B. E. M. Steenkamp, 2002. Identifying spatial
segments in international markets. Marketing Science, 21(2):160-177.
Vermunt, J. K., 2001. The use of restricted latent class models for defining and
testing nonparametric and parametric item response theory models. Applied
Psychological Measurement, 25(3):283-294.
Vermunt, J. K., 2004. An EM algorithm for the estimation of parametric and non-
parametric hierarchical nonlinear models. Statistica Neerlandica, 58(2):220-
233.
Vmw Analytic Service, 2011. Economy of the federal states for year 2010 &
population for year 2011. (Available at www.unidow.com).
B IBLIOGRAPHY 139
Wedel, M., F. Ter Hofstede, and J.-B. E. M. Steenkamp, 1998. Mixture model
analysis of complex samples. Journal of Classification, 15(2):225-244.
Wind, Y., 1978. Issues and advances in segmentation research. Journal of Market-
ing Research, 15(3), 317-337.
Zhou, K. Z., C. Su, and Y. Bao, 2002. A paradox of price-quality and market
efficiency: A comparative study of the US and China markets. International
Journal of Research in Marketing, 19(4):349-365.
140 V R ESEARCH S HOPPING
Abstract
Based on two studies, this article reveals the underlying mechanisms that deter-
mine research shopping, described as switching channels between searching and
purchasing a product or service. Findings indicate that the status quo bias theory
is valuable to determine consumers biased perceptions as well as biased behavior
in terms of the reluctance to research shopping. These two types are investigated
via the channels employed for searching and overall evaluations of the search
phase. In doing so, this article contributes to the multichannel literature by sub-
stantiating knowledge about research shopping and adding to the knowledge about
the difference among multichannel and single-channel consumers. Additionally,
it strengthens the status quo bias theory as theoretical underpinning in marketing
literature.6
Keywords: Research shopping, Status quo bias theory, Biased channel perceptions,
Biased channel choice, Multichannel search behavior
6 T. Schlager. This paper is in preparation for submission to the "International Journal of Research
in Marketing".
1 Introduction 141
1 Introduction
More and more consumers are taking advantage of the broad variety of available
channels. Accordingly, an increasing number of consumers uses different channels
to search and purchase products or services. For instance, a recent study of Google
(2009) illustrates that 40 percent to 60 percent of German consumers, depending
on the product, switch channels between the search and the purchase. In academic
research, this phenomenon, described as "the propensity of consumers to research
the product in one channel [. . . ], and then purchase it through another channel",
has been dubbed research shopping (Verhoef et al., 2007, p. 129).
shopping behavior and research shopping in particular are loudening (Chiu et al.,
2011; Kopalle et al., 2009; Pauwels et al., 2011; Pentina et al., 2009).
3. The article adds knowledge regarding the rarely applied status quo bias
theory by projecting its thoughts to a marketing setting.
2 Conceptual model development 143
Consumers have multiple choices regarding what channel to use in each phase of
the buying process. In case the consumer switches channels between the search
and purchase phase research shopping is present. Now assume that, contrary to
Verhoef et al. (2007), multiple channels can be used for searching. For instance,
one channel is central to information acquisition; however, a second channel is
used to complement the information at hand. Thus, considering this extension
to the article of Verhoef et al., two search situations are given, one for a single
channel search and the extension for a multichannel search. For clarity, let the
most important search channel be the primary search channel and the concurrent
channel be the alternative search channel.
2003). Thus, it is rather the relative evaluation of the utilities of the primary and
alternative search channel for making the purchase that is decisive for research
shopping.
A significant part of the literature has shown instances where behavior deviates
from what traditional economic theory predicts (e.g., Kahneman et al., 1991).
More specifically, the assumption that the expected utility is the sole determinant
of consumer behavior does not necessarily hold true (Hershey et al., 1982). A
notable exception is presented by the status quo bias theory (e.g., Samuelson and
Zeckhauser, 1988), which is described as the "propensity of a decision-maker to
choose an option because it is the status quo, the default option" (Dean, 2008, p.
4). From this perspective, consumers display a higher tendency to remain with the
primary search channel throughout the purchase process. They will only switch
channels to make a purchase when the alternative search channel is perceived
better by far as in this case consumers have sufficient motive for research shopping
(Falk et al., 2007; Gal, 2006). This outlines the first facet suggested for the status
quo bias: Consumers are reluctant to research shopping although higher utility is
derived from purchasing in the alternative search channel (biased behavior).
Besides the directly biased behavior, consumers may also be biased in their
perceptions. For instance, Falk et al. (2007) describe the status quo bias as
outcome of systematically biased perceptions of a new self-service channel since
these drive the probability to keep the status quo channel. Thus, it is suggested
that consumers perceive higher utility of the primary search channel for making
the purchase solely due to representing the status quo, while perceiving less utility
of alternative search channel. This describes the second type of the status quo bias:
Consumers indirectly avoid research shopping since the perceptions of the primary
and alternative search channel are systematically altered (biased perception).
In sum, this paper suggests two mechanisms of how the status quo bias theory
3 Hypotheses development 145
applies to research shopping. First, consumers have the propensity to remain with
the search channel for purchasing, even if the alternative channel is evaluated as
superior. Second, the perceptions of the channels are systematically influenced
by the status quo option; the primary search channel is evaluated more favorably
for purchasing, whereas the alternative channel is evaluated worse resulting in a
lower probability of research shopping (see Figure 10).
Figure 10: Conceptual Model: The two Potential Status Quo Biases in Research
Shopping.
3 Hypotheses development
making the purchase. More precisely, their status quo is the use of both channels,
however, they need to make a choice for making the purchase. Accordingly, two
situations with different utilities derived from research shopping are identified
and by comparing both consumer types, the status quo bias is isolated and its
magnitude determined.
In the first situation, a consumer relies on one channel for searching which
is suggested to be equivalent to the status quo. As Verhoef et al. (2007) and
Konus et al. (2008) suggest, despite several attributes determining the utility
across several phases of the purchasing process, some are phase-specific. For
instance, attributes, such as the ease of comparing different options are highly
valued when searching, while others, such as the effort for buying are rather related
to purchasing. Although the consumer cannot fully foresee the utility of buying
with the primary search channel, even less can be said about the alternative search
channel. Correspondingly, when searching in one channel the consumer knows
less about the specific utility of the alternative search channel for purchasing as the
consumer did not experience the alternative search channel in the search phase.
a vague outcome as the status quo and a clearly defined outcome as the alternative.
As most participants retained their current status, it was concluded that the bias
decreases the perceived uncertainty towards the status quo.
The above argumentation implies that the difference in the perceived uncer-
tainty between the primary and the alternative search channel will be large for
consumers who only use one search channel. Accordingly, the propensity for us-
ing the primary search channel for purchasing increases (Bewley, 2002; Ortoleva,
2008; Samuelson and Zeckhauser, 1988). Even when the perceived utility of the
alternative search channel is higher for purchasing than the utility of the primary
search channel, the primary channel is given a higher preference. Consequently, it
is proposed that consumers tend to avoid research shopping when they have only
used one channel for searching favoring the status quo bias.
H 1a. The difference in perceived attributes of the alternative and the primary
search channel for making the purchase determines the probability of research
shopping.
H 1b. Using multiple channels in the search process increases the probability of
research shopping.
After reflecting the peculiarities of both situations, the status quo bias calls
for further investigation especially concerning how the bias fully unfolds its
magnitude. Specifically, literature has argued that consumers who search in one
channel need to overcome the barrier induced by a status quo bias (Falk et al., 2007;
Gal, 2006), whereas consumers that search in multiple channels are supposed
to switch the primary search channel for making the purchase when perceiving
the alternative channel superior to the primary search channel. Registering both
situations, it is suggested that the difference in the utility of research shopping
between using multiple and single channels during the search phase is largest
when consumers expect a slightly higher utility from the alternative search channel
than from the primary search channel; multichannel search consumers would
immediately switch when expecting the alternative channel to be superior for
buying. Accordingly, the difference between using multiple channels or a single
channel in the search phase will be smaller at the extreme of the relative evaluation.
Further, it is suggested that the curve will peak beyond the point where the
alternative search channel is perceived as equal to the primary search channel,
since both remain with the primary search channel when perceiving it better for
making the purchase. As a result, H 1c-d are suggested:
H 1c. The curve reflecting the difference in the probabilities of research shopping
between multichannel and single-channel consumers is inversely U-shaped.
H 1d. The difference in probability curve will peak beyond the point where
equally perceiving both channels for making the purchase.
3.2 Biased perceptions of channels for purchasing 149
In the conceptual model altered perceptions of the primary and the alternative
search channel were suggested to present the second type of status quo bias. To
shed light on this, the single channel search situation, which is suggested to be
necessary for a status quo bias, is further investigated. In this course, literature
suggests that once a status quo is defined, it is taken as a reference (Anderson,
2003; Falk et al., 2007). That is, every upcoming alternative is evaluated in the
light of the status quo. When the current state is perceived as particularly favorable,
the absolute magnitude of losses is larger since the well-perceived state is at stake
(e.g., Anderson, 2003; Falk et al., 2007; Yen and Chuang, 2008). For instance,
Yen and Chuang (2008) argued that "happy individuals who feel confident in
following a business as usual procedure should be more likely to select the status
quo option" (p. 524). In a series of experiments, they found support for this,
while negative emotions mitigated keeping the status quo. Martin et al. (1993)
supports this notion showing that consumers who report negative evaluations
rather rely on systematic processing, while positive evaluations enhance heuristic
information processing. Hence, when negatively evaluating the primary search
channel, consumers more thoroughly re-evaluate both channels for making the
purchase, whereas a positive evaluation may even exclude the alternative search
channel from being considered for purchasing. Correspondingly, it is suggested
that favorable evaluations of the search phase directly enhance the propensity to
maintain the status quo, while unfavorable evaluations increase the probability of
research shopping.
point that provokes the status quo bias. Hence, the following hypothesis can be
established:
H 2a. Satisfaction with the search channel decreases the probability of research
shopping.
Besides these initial suggestions, the status quo bias theory explicitly suggests
irrational elements (e.g., Kahneman et al., 1991). In line, unrelatedly altered per-
ceptions are one central indication for the presence of a status quo bias (Baron and
Ritov, 1994). These are present when the consumers perceptions of an alternative
channels attributes are influenced by the overall evaluation with the primary chan-
nel for searching. A trigger of the status quo that alters consumers perceptions is
loss aversion. Loss aversion delineates that consumers overemphasize potential
losses compared to gains (e.g., Bostrom and Ord, 2006; Kahneman et al., 1991;
Rubaltelli et al., 2005; Yen and Chuang, 2008). As a result of loss aversion,
subjective certainty with the status quo increases (Yen and Chuang, 2008). In the
context of research shopping, this implies that loss aversion concerning the cur-
rent state decreases, resulting in better-perceived attributes of the primary search
channel. For the alternative search channel, however, loss aversion leads to even
worse perceived attributes for the alternative channel. In line, Falk et al. (2007)
suggested that the intention to adapt an unknown channel is primarily influenced
by the evaluation of currently used channels. Studying offline-banking customers
they discovered a negative influence of satisfaction with the offline channel on the
perceptions of the newly introduced self-service channel. Perceptions were found
to be biased which in turn lowered the usage intentions for the self-service channel.
In particular, it is suggested that consumers tend to systematically perceive the
alternative search channels attributes for purchasing as worse, while they more
positively evaluate the primary search channels attributes.
To sum up, using a well-perceived current state has been noticed as a decisive
facet in provoking a status quo bias since larger potential losses are perceived (e.g.,
Falk et al., 2007; Yen and Chuang, 2008). Correspondingly, it is proposed that
3.2 Biased perceptions of channels for purchasing 151
both channels attributes are misperceived in favor of the primary search channel
when a clearly defined status quo exists (e.g., Kim and Kankanhalli, 2009). As
such, these evaluations lower the tendency to switch the channel for making the
purchase. It is noteworthy that the suggested mechanism is substantially different
from the lock-in effect, which specifies the superiority of a specific combination
of channels (e.g., Verhoef et al., 2007) or switching costs (e.g., Shankar et al.,
2003) since both mechanisms would not imply misperceptions of the alternative
channel (Falk et al., 2007). Following three hypotheses are established:
H 2b. The perceived attributes of the search channels for purchasing partially
mediate the influence of satisfaction with the search channel on the probability
of research shopping.
H 2c. Satisfaction with the search channel negatively influences the perceived
attributes of the alternative channel for purchasing.
H 2d. Satisfaction with the search channel positively influences the perceived
attributes of the primary channel for purchasing.
To sum up, two types of status quo bias are suggested. First, biased behavior is
investigated by comparing the combination of channels used for searching. In case
two channels are used, it is suggested that consumers display unbiased behavior.
Contrary, consumers are suggested to be reluctant to research shopping when only
one channel was applied for searching, which is in accordance with the status quo
bias theory (H1a-c). Second, consumers perceptions of the search channels used
for purchasing are biased when being satisfied with the search phase (H2a-d). The
central hypotheses are summarized in Figure 11.
152 V R ESEARCH S HOPPING
Notes:
Attalternative are the perceived attributes of the alternative search channel
for making the purchase.
Att primary are the perceived attributes of the primary search channel
for making the purchase.
Figure 11: The Central Hypotheses concerning the Effects of the Status Quo
Bias Theory on Research Shopping.
For the large-scale study, a survey was developed asking participants about
their buying processes regarding financial services. Draft versions were thor-
oughly discussed with industry executives in the participating countries to gain
insights regarding country-specific issues. Additionally, a pretest with 589 U.S.
respondents was conducted. Following the participants comments, scales and
items were adapted. After the survey was sent out for translation into thirteen
languages, it was discussed at least twice with contact persons of each country to
determine whether fine-grained culture or language-specific differences posed a
problem (Burgess and Steenkamp, 2006). To distribute the questionnaire, panels
representative for each participating country were used. The questionnaire was
accompanied with a cover page that explained the purpose of the study. After
initial questions, the channels were described to facilitate respondents in correctly
classifying search and purchase channels. The procedure provided 16,276 useful
questionnaires from five continents (Asia, Australia, Europe, North America, and
South America). Sample characteristics are distributed as follows: Mean age
was 39.4 (S.D. = 11.66); 48.56 percent were female and 51.54 percent male;
most respondents graduated from a university (58.6 %), a third (34.4 %) had a
high-school degree, and the rest (7.0 %) had not completed any degree. In sum,
9,760 persons used the offline channel for searching as primary search channel,
while 6,516 chose the online option. However, 11,611 participants took the offline
choice for buying. Only 4,665 made their purchase via an online channel.
Measurement
Consumers were directly asked to specify the channels they used for searching,
as well as their primary search channel. To validate that the primary channel was
dominantly used, consumers were required to disclose the amount of time spent
in each channel. If a consumer spent twice as much time in the primary search
154 V R ESEARCH S HOPPING
Confirmatory factor analysis (CFA), based on offline and online channels, was
employed to condense the items to factors and to assess their unidimensionality,
as well as their validity. The results were verified by the commonly used fit-
indices except the c 2 which would be inflated due to the extensive sample size
(Jreskog, 1969). They indicated that the data fit the model well: The standardized
loadings were high and significant (see Appendix A), ranging from .54 to .85 (a
< .001 level), and in conjunction with a battery of fit-indices the model fit was
4.2 Study 1: Analysis 155
Search channel (off-, online) 2.45 .4 .49 -.04 *** .09 *** .04 *** .09 *** .25 *** 1
Search channels used 2.01 .53 .5 -.02 * .01 0 -.02 * .08 *** -.03 ** 1
Time spent for search (%) 5.92 51.43 19.17 .03 *** -.01 -.01 .03 *** -.12 *** -.02 ** -.05 *** 1
Buying Usefulness 2.37 -.76 1.81 -.09 *** .04 *** -.01 .04 *** .56 *** .47 *** .05 *** -.13 *** 1
Buying Risk 2.19 1.13 2.26 .09 *** -.04 *** 0 -.04 *** -.51 *** -.51 *** -.05 *** .1 *** -.69 *** 1
Buying Ease 1.16 -.32 2.1 -.06 *** -.01 0 -.04 *** .26 *** -.04 *** .08 *** -.09 *** .29 *** -.19 *** 1
Notes:
V R ESEARCH S HOPPING
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001.
Correlations of orthogonalized factors shown (as used in regression).
The probability of the binary outcome (i.e., research shopping) was estimated
using a random effect logit model. Let rsic be a binary variable, specifying if
research shopping is present, rsic = 1, or not, rsic = 0, for a consumer i using a set
of search channels c (c = 1 when using a single search channel, otherwise c = 2).
The relative utility related to research shopping is specified as latent variable, rsic .
This is denoted as follows:
8
< 1 if rsic 0
rsic , i = 1 , ..., n (4)
: 0 if rsic < 0
Four groups of variables reflect and isolate the patterns described by the con-
ceptual model: Consumer, product as well as search phase-specific characteristics,
and the relative evaluation of the perceived attributes of the search channels for
making the purchase. We can denote the relationship between these four groups
and the latent variable as
where
158 V R ESEARCH S HOPPING
polynomials included). All factors show high significance for explaining research
shopping. Specifically, buying perceived usefulness yields the highest impact
among the factors (b = .453, p < .001). Similarly, perceived risk of buying sig-
nificantly decreases the probability of switching channels between searching and
buying (b = -.212, p < .001). The factor with the least influence is the perceived
ease of buying of a channel (b = .127, p < .001). In sum, hypothesis H1a was
broadly supported.
To interpret the effect of using multiple against a single channel for searching
(H1b-H1d) simulation was used (Zelner, 2009). In particular, the posterior means
of the predicted probabilities of research shopping were simulated for both situa-
tions varying the three explanatory variables (perceived usefulness, perceived risk,
and perceived ease of use), all of the other variables held at their observed values
(Van Everdingen et al., 2011). In a similar vein, the difference in the probability
of research shopping between both situations was checked. The results confirm
that using multiple channels within the search phase positively influences the
probability of research shopping, offering support for H1b. For buying perceived
usefulness, perceived risk, and perceived ease of use (differentiated by offline
and online search) they are displayed in Figure 12, which is equivalent to the
visualization of Van Nierop et al. (2008). When searching offline, no inverse
U-shaped curve was found for perceived usefulness and perceived risk since the
curve did not decrease at the extremes of the scale. However, for perceived ease
of use the curve yielded an inverse U-shaped form. For searching online the in-
versely U-shaped was supported for all three explanatory variables. The effect of
using two channels, instead of one, has the highest influence when the alternative
search channel is evaluated as slightly better than the primary search channel.
When the attributes of the alternative channel for buying are perceived better by
far, consumers switch channels for purchasing, regardless of the applied search
channels. Thus, H1c and H1d were partially confirmed.
160
Model (1) Model (2) Model (3) Model (4) Model (5) Full model Single search Multi search
Covariates only No nonlinearities Polynomials usefulness Polynomials risk Polynomials ease All polynomials All polynomials All polynomials
b p SE b p SE b p SE b p SE b p SE b p SE b p SE b p SE
Age -.000 (.002) .002 (.002) .002 (.002) .002 (.002) .002 (.002) .002 (.002) .001 (.003) .003 (.003)
Gender -.053 (.044) -.039 (.045) -.042 (.045) -.036 (.045) -.046 (.045) -.044 (.045) -.118 (.073) .018 (.058)
Experience -.088 + (.046) -.089 + (.047) -.090 + (.047) -.089 + (.047) -.093 * (.047) -.094 * (.048) -.149 * (.076) -.045 (.062)
Product complexity -.132 ** (.046) -.135 ** (.047) -.135 ** (.047) -.134 ** (.047) -.133 ** (.047) -.130 ** (.048) -.235 ** (.076) -.047 (.062)
Price .227 *** (.012) .073 *** (.014) .075 *** (.014) .074 *** (.014) .072 *** (.014) .075 *** (.014) .082 *** (.022) .069 *** (.018)
Searchchannel (off-, online) 2.003 *** (.048) 1.556 *** (.054) 1.537 *** (.055) 1.576 *** (.055) 1.551 *** (.054) 1.563 *** (.055) 1.290 *** (.092) 1.708 *** (.069)
Time spent in search (%) -.014 *** (.001) -.013 *** (.001) -.013 *** (.001) -.013 *** (.001) -.013 *** (.001) -.013 *** (.001) -.009 *** (.002) -.016 *** (.002)
DBuyingUse f ulness .488 *** (.035) .461 *** (.039) .492 *** (.035) .488 *** (.035) .453 *** (.041) .391 *** (.064) .496 *** (.053)
DBuyingRisk -.231 *** (.033) -.231 *** (.033) -.223 *** (.037) -.232 *** (.033) -.212 *** (.038) -.282 *** (.064) -.174 *** (.048)
DBuyingEase .112 *** (.026) .112 *** (.026) .114 *** (.026) .121 *** (.027) .127 *** (.027) .186 *** (.042) .082 * (.036)
DBuyinguse f ulness2 .038 (.030) .061 + (.034) .076 (.054) .046 (.044)
DBuyingUse f ulness3 -.081 ** (.027) -.095 *** (.028) -.083 + (.045) -.105 ** (.036)
DBuyingRisk2 .012 (.028) .031 (.031) -.005 (.052) .047 (.039)
DBuyingRisk3 -.053 * (.025) -.075 ** (.026) -.084 * (.042) -.069 * (.034)
DBuyingEase2 -.107 *** (.026) -.122 *** (.027) -.123 ** (.043) -.109 ** (.036)
DBuyingEase3 -.033 (.026) -.028 (.026) -.047 (.041) -.023 (.035)
Constant -1.545 *** (.346) -1.729 *** (.340) -1.723 *** (.341) -1.733 *** (.340) -1.753 *** (.338) -1.744 *** (.338) -2.030 *** (.171) -1.355 *** (.141)
Intercept -.761 *** (.502) -.785 *** (.503) -.783 *** (.503) -.785 *** (.503) -.792 *** (.502) -.792 *** (.503)
a a
N 16,276 16,276 16,276 16,276 16,276 16,276 7,622 8,654
AIC 13199.198 12701.311 12696.498 12698.763 12684.310 12669.999 5148.946 7503.000
c2 2474.031 2650.921 2684.898 2646.711 2655.532 2703.289 862.489 1861.979
LL -6590.599 -6338.655 -6334.249 -6335.382 -6328.155 -6317.000 -2556.473 -3733.500
V R ESEARCH S HOPPING
LL comparison -6804.952 -6534.276 -6530.177 -6530.413 -6520.447 -6508.402 a a
LR test base model - 508.89*** 8.81* 6.55* 21.00*** 43.31*** a a
Notes:
Unstandardized estimates shown.
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001.
a) No random intercepts used since the sample was split according to the specific search channels used.
Notes:
The blue line indicates the probability of research shopping when using multiple channels for
for searching.
The green line indicates the same values when using one channel for searching.
The red solid line indicates the predicted change in the probability of research shopping
The red dotted lines indicate the boundaries of the 95 % confidence intervals.
The explanatory variable is varied along the horizontal axis holding the nonfocal variables at
at their observed values.
Study 2 implemented a 2 (offline versus online) x 2 (high risk versus low risk)
x 2 (high usefulness versus low usefulness) between-subjects experimental design;
in sum eight conditions were developed. The treatments relate to the factors
of the survey and manipulate the level of satisfaction as in previous research
on satisfaction (e.g., Homburg et al., 2005). All manipulations were tested in
advance. To manipulate the offline search channel, video screenings with a tied
agent were recorded and shown to the participants in a separate room to guarantee
a homogeneous treatment. For the online treatment, four different websites were
designed. Table 9 offers a description of each manipulation.
Measurement
Satisfaction with the search channel was assessed using a three-item construct,
previously used by Voss et al. (2010). As the scale has not yet been applied in
the channel context, it was slightly adapted. Consumer satisfaction forms the
independent variable and directly influences the perceptions of the search channels
attributes for purchasing and research shopping. To conceptualize the mediating
constructs, the same items as in Study 1 were used; however, a seven-point scale
was applied. Research shopping was measured by directly asking whether the
participant would also purchase in the search channel.
Factors Outcome
Study 2: Results
To test the hypotheses 2a-d, linear (for the relationships between satisfaction
and the perceptions), as well as logistic regression (for the probabilistic determina-
tion of research shopping) analysis were applied. As suggested by H2a and H2b,
4.4 Study 2: Analysis 165
it was tested for the potential mediation effect of satisfaction with the search phase
by the perceived attributes of the channels for buying. Again, the relative evalua-
tion of channel attributes was used. In terms of the dichotomous outcome variable
research shopping, the method proposed by MacKinnon and Dwyer (1993) and
further discussed by Kenny (2008) was applied. Specifically, each coefficient
was multiplied by the predictors standard deviation and subsequently divided by
the outcome variables standard deviation. This process aims at guaranteeing the
coefficients comparability. Next, direct and indirect effects were bootstrapped.
The results from the estimation largely support the established hypotheses.
V R ESEARCH S HOPPING
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001.
BU = buying perceived usefulness.
BRISK = buying perceived risk.
BE = buying perceived ease.
SSAT = search channel satisfaction.
RS = Research shopping (1/0).
The central hypotheses regarding satisfaction with the search channel (H2c
and H2d) proposed that satisfaction with the search channel is positively related
to the perceptions of the primary search channel for purchasing, and negatively
related to the perceptions of the alternative search channel for purchasing. In
the overall model all hypotheses can be confirmed. Satisfaction significantly
influences perceived usefulness (b = .399; p < .001), perceived risk (b = -.329;
p < .001), and perceived ease of use (b = .087; p = .008) of the primary search
channel for purchasing. Similarly, satisfaction with the search channel unfavorably
influences the perceptions of the alternative search channel for purchasing. The
regression estimates were highly significant and meaningful; satisfaction with the
search channel unfavorably influences perceived usefulness (b = -.101; p = .002),
perceived risk (b = .139; p < .001), and perceived ease of use (b = -.114; p < .001)
of the alternative search channel. Finally, it was checked for differences between
searching offline and online as Study 1 indicated that they exist.
The results for searching offline clearly pointed out that satisfaction leads to
systematically worse perceptions of the alternative channels attributes (perceived
usefulness: b = -.072; p = .008, perceived risk: b = .178; p < .001, perceived ease
of use: b = -.130; p = .003), while perceiving the primary channels attributes more
favorable (perceived usefulness: b = .430; p < .001, perceived risk: b = -.347; p <
.001, perceived ease of use: b = .072; p = .087). Contrary to this, the hypotheses
for searching online were not fully confirmed. Satisfaction with the online search
favorably influences perceived usefulness (b = .340; p < .001), perceived risk (b
= -.266; p < .001), and perceived ease of use (b = .136; p = .012) of the primary
search channel for purchasing. However, only one of the hypothesized negative
links between satisfaction with the search channel and the perceptions of the
alternative channel was found, namely for the influence of satisfaction on the
perceived ease of use of the alternative channel (b = -.117; p = .027). All other
hypotheses were rejected (perceived usefulness: b = -.011; p = .810; perceived
risk: b = -.004; p = .925). To sum up, the influence of satisfaction is particularly
168 V R ESEARCH S HOPPING
salient when searching offline. However, the proposed effect remains rather silent
when searching online.
5 Discussion
in line with Verhoef et al.s (2007) findings that online consumers are more prone
to research shopping. In sum, although the hypotheses were not fully supported, it
is suggested that consumers who search online were less susceptible to a status
quo bias.
be considered to provide a complete picture of how the status quo bias theory is
applicable to research shopping.
6 Managerial implications
Assuming that the biases found in this study cannot be directly used to advise
companies in refining their multichannel strategies would be misleading. Research
has provided substantial evidence that multichannel consumers are more profitable
than consumers who use a single channel (e.g., Kumar and Venkatesan, 2005;
Venkatesan et al., 2007). To persuade single channel consumers to become (more
profitable) multichannel consumers, the status quo bias must be overcome. The
necessity of a well-integrated and orchestrated system of channels is suggested
(Chatterjee, 2010; Neslin et al., 2006; Reynolds et al., 2006). For instance,
practitioners could provide online terminals in their brick-and-mortar stores to
reduce uncertainty toward the commonly more cost-efficient online channels
(Campbell and Frei, 2010; Hitt and Frei, 2002). Accordingly, direct links to
offline channels on company websites may reduce existing uncertainties when
buying offline. Both options enhance the utility that is derived from buying within
the same companys channels by diminishing a potential bias (Gal, 2006) to
switch the channel. In line with recent research (Ansari et al., 2008; Thomas
6 Managerial implications 171
Finally, in line with Valenzuela et al. (2008), it was suggested that the presence
of a status quo bias may also result in higher levels of satisfaction with the buying
process. Consumers might not feel being forced to make a trade-off between what
channel to employ because they are biased toward the search channel; therefore,
not being status quo-biased may be negatively related to the satisfaction of the
overall buying process. Thus, when striving to make consumers search and
purchase with the same channel, practitioners can directly use this bias to enhance
overall consumer satisfaction with the buying process.
As with all studies, the current investigation has limitations that can guide future
research. First, this article focuses on online and offline channels to provide
initial evidence on a potential status quo bias. Previous literature has investigated
a broader variety of channels (i.e. catalogs, Verhoef et al., 2007) using other
approaches and theoretical underpinnings. While the plurality of channels has
much increased, few resonances can be found in research. Thus, considering the
status quo bias for other types of channels constitutes an opportunity for further
research.
is suggested that the status quo bias theory holds important implications for this
phenomenon. In line, we expect interesting findings from using the status quo
bias theory for further channel-related phenomena.
In line, deciphering the effects of the status quo bias theory is an example for
applying alternative theories to a phenomenon that has primarily been considered
from similar theoretical perspectives, such as the theory of reasoned action or
planned behavior (e.g., Pookulangara and Natesan, 2010; Verhoef et al., 2007).
This allows for generating novel theoretical insights since both aforementioned
theories are in line with economics theory. It is worthwhile to note that not
all hypotheses were confirmed, thus, further theories may be helpful to explore
consumer channel behavior and provide additional insights into research shopping.
For instance, combining switching cost theoretical considerations (e.g., Burnham
174 V R ESEARCH S HOPPING
et al., 2003; Shankar et al., 2011) with the status quo bias theory may provide
information on the boundaries of the selected theoretical perspective. Certainly,
also theories from other disciplines, such as consumer psychology, have the
potential to further complement the current findings.
A highly attractive avenue for future research is to examine the status quo
bias in the combination of research shopping and switching companies, which
is dubbed cross channel shopping (e.g., Brynjolfsson et al., 2009; Chiu et al.,
2011). This is certainly attractive as profound knowledge regarding the correlation
between using different channels and companies when searching and purchasing
a product or service is still missing (e.g., Chiu et al., 2011; Nunes and Cespedes,
2003).
Finally, compared to studies that use overall evaluations of risk and usefulness
for determining channel choice (e.g., Montoya-Weiss et al., 2003; Verhoef et al.,
2007), consumers were directly asked for their perceptions of channel attributes.
Certainly, one limitation of this procedure is its difficulty in terms of the confir-
matory factor analysis. Although the factors employed were largely supported
by both analyses, items needed to be dropped because of multi-dimensionality.
Moreover, one factor out of twelve did not reach an average variance extracted
of .5, which did not alter the basic mechanism since the effects in the follow-up
analyses were deemed substantial.
Appendices
B Graphical abstract
B IBLIOGRAPHY 177
Bibliography
Baron, J. and I. Ritov, 1994. Reference points and omission bias. Organizational
Behavior and Human Decision Processes, 59(3):475-498.
Bostrom, N. and T. Ord, 2006. The reversal test: Eliminating status quo bias in
applied ethics. Ethics, 116(4):656-679.
Campbell, D. and F. Frei, 2010. Cost structure, customer profitability, and reten-
tion implications of self-service distribution channels: Evidence from customer
behavior in an online banking channel. Management Science, 56(1):4-24.
Chernev, A., 2004. Goal orientation and consumer preference for the status quo.
Journal of Consumer Research, 31(3):557-565.
Chiu, H.-C., Y.-C. Hsieh, J. Roan, K.-J. Tseng, and J.-K. Hsieh, 2011. The
challenge for multichannel services: Cross-channel free-riding behavior. Elec-
tronic Commerce Research and Applications, 10(2):268-277.
Ciccone, A., 2004. Resistance to Reform: Status quo bias in the presence
of individual-specific uncertainty: Comment. American Economic Review,
94(3):785-795.
Dean, M., 2008. Status quo bias in large and small choice sets, mimeo.
Gal, D., 2006. A psychological law of inertia and the illusion of loss aversion.
Judgment and Decision Making, 1(1):23-32.
Hartman, R. S., M. J. Doane, and C. K. Woo, 1991. Consumer rationality and the
status quo. Quarterly Journal of Economics, 106(1):141-162.
Hawkins, D. I., G. Albaum, and R. Best, 1974. Stapel scale or semantic differential
in marketing research? Journal of Marketing Research, 11(3):318-322.
Hu, L. and P. M. Bentler, 1999. Cutoff criteria for fit indexes in covariance
structure analysis: Conventional criteria versus new alternatives. Structural
Equation Modeling, 6(1):1-55.
Kumar, V. and R. Venkatesan, 2005. Who are the multichannel shoppers and how
do they perform? Correlates of multichannel shopping behavior. Journal of
Interactive Marketing, 19(2):44-62.
Ortoleva, P., 2008. Status quo bias, ambiguity and rational choice. (Available at
http://www.gtcenter.org).
Ritov, I. and J. Baron, 1992. Status-quo and omission biases. Journal of Risk and
Uncertainty, 5(1):49-61.
Schaefer, M., 2011. Winning over the empowered consumer - Why trust matters.
IBM Institute for Business Value, Somers, NY.
(Available at http://public.dhe.ibm.com).
Spranca, M., E. Minsk, and J. Baron, 1991. Omission and commission in judgment
and choice. Journal of Experimental Social Psychology, 27(1):76-105.
Van Nierop, E., D. Fok, and P. H. Franses, 2008. Interaction between shelf layout
and marketing effectiveness and its impact on optimizing shelf arrangements.
Marketing Science, 27(6):1065-1082.
Voss, G. B., A. Godfrey, and K. Seiders, 2010. How complementarity and substi-
B IBLIOGRAPHY 185
Yahoo! Inc. and OMD, 2006. Long & winding road: The route to the cash register.
(Available at http://files.shareholder.com/).
Zelner, B. A., 2009. Using simulation to interpret results from logit, probit, and
other nonlinear models. Strategic Management Journal, 30(12):1335-1348.
Curriculum Vit
Personal Information
Education
Work Experience