how to enable cloud success? unveiling the ... · how to enable cloud success? unveiling the...
TRANSCRIPT
1
HOW TO ENABLE CLOUD SUCCESS?
UNVEILING THE MULTIDIMENSIONAL NATURE OF
CLOUD SERVICE VIABILITY
Abstract
Providers of consumer cloud services face enormous challenges in developing a
sustainable market position in this highly competitive market. Emergent trends like
consumerization lead to high growth rates and extend the reach of these services
far into the enterprise sphere. Using a freemium model, many providers focus on
establishing a large customer base quickly. Unfortunately, they often lack a strategy
to generate revenue streams in the long run. Based on a sample of 596 actual
users, our study examines how consumer cloud services can become viable, i.e.,
being self-sustainable on the basis of the user base and revenue streams they
generate. The results indicate that focusing on a single performance indicator is not
sufficient to understand the viability of cloud services. We conceptually differentiate
between different types of willingness to pay that have been used synonymously in
previous studies and provide empirical evidence that their drivers differ
fundamentally. Thereby, we establish the multidimensional nature of cloud service
viability as a promising perspective to study cloud service scenarios. The key
findings are used to derive specific recommendations for three generic strategies
Authors / Viability of Cloud Services
2
that cloud providers can apply to become viable in their competitive market
environment.
Keywords: Cloud computing, viability, online consumer behavior, cloud services,
willingness to pay, freemium, business models, word-of-mouth.
Authors / Viability of Cloud Services
3
Introduction
The consumer cloud service market is highly competitive and IT providers face
enormous challenges in positioning their services (Rossbach & Welz, 2011). In
contrast to enterprise services, many providers focus on establishing a large
customer base quickly. In the long run however, these services often lack a strategy
to generate sufficient revenue streams (Needleman & Loten, 2012). Others charge
users immediately or after a trial phase, but don’t manage to generate enough
growth. The difficulty of cloud services to become viable – i.e., being self-
sustainable on the basis of both the user network and revenues it generates – shall
be illustrated by a few cases. Successful examples for cloud services are
Freshbooks, a cloud-based accounting service, Mailchimp, a cloud-based marketing
service, or Evernote, a service for note taking and archiving. All of those offer limited
free versions of their services but charge users for advanced features or
requirements, also referred to as freemium model (Teece, 2010). The revenue
model has also been widely adopted by a variety of social networking or news
platforms (Niculescu & Wu, 2014). While growing fast, all these services managed
to continuously increase their number of paying users (e.g., Chestnut, 2010).
However, the transformation of free users into paying consumers can also fail (S. S.
Kim & Son, 2009). Experts estimate that - on average - only 2% of the users pay for
freemium-based cloud services (Needleman & Loten, 2012). Cases where a focus
on creating a large user base (instead of focusing on generating revenues) led to
problems are manifold. For instance, Chargify LLC, a provider of billing-
management software, was also motivated to use the freemium model by the
Authors / Viability of Cloud Services
4
seemingly low marginal costs of a cloud-based service. When Chargify started out in
2009, they were very successful in generating users (Needleman & Loten, 2012).
However, the company was soon on the path to bankruptcy because users never
became paying customers. They decided to change the focus of their strategy from
generating users to preliminarily generating revenues. Accordingly, they put up a
paywall for all users at the cost of lower growth rates but became self-sustainable
with more than 900 paying customers in 2012 (Needleman & Loten, 2012). A similar
turn was necessary for Ning, a platform to create social websites (Rao, 2010). The
firm also discontinued the free service and successfully introduced a small fee for all
users. In contrast, other companies such as iOctocat, a GitHub application, were
initially driven by revenue generation, but had to switch to a freemium model to grow
faster (Reimann, 2013). In summary, the examples highlight, that both, a focus on
growth and a focus on revenues, can have dramatic consequences for cloud service
providers. Thus, we suggest that a variety of performance indicators and their
interrelationships must be considered to understand the rise or stagnancy of cloud
services. In this article, we address theoretically and empirically how and why cloud
services become viable, i.e. become self-sustainable on the basis of the user base
and revenues they generate.
A service is viable if it is “capable of succeeding” (Merriam-Webster dictionary).
Therefore, viability does not refer to the long term success of a business. Rather, it
can be described as a customer-oriented configuration that provides the capabilities
for succeeding. If a service is supposed to be successful in the long run, it has to be
viable. However, if it is not viable, it can never be successful. Hence, viability is a
necessary building block for successful cloud services. In this study, we examine
Authors / Viability of Cloud Services
5
five customer-related key performance indicators (KPIs) that measure the viability of
a cloud service, namely customer satisfaction, loyalty, word-of-mouth, willingness to
pay for retention (WTPR) and willingness to pay for an upgrade (WTPU) with
respect to their main drivers and interrelationships. As we discuss in the following,
these KPIs indicate to what extent a cloud service is able to build and retain a solid
user base and transform users into paying customers.
The study focusses on consumer IT services for several reasons. While enterprise
services have been addressed by many researchers in IS, technological
advancements have driven the individuation of IS (Baskerville, 2011). This trend is
especially prevalent for cloud services, that eliminate an up-front commitment by the
users, allowing them to start small and increase or reduce computing resources as
needed (Armbrust et al., 2010). Two further prevalent developments drive the need
to understand the viability for consumer cloud services beyond the case of individual
consumers themselves: First, more and more gadgets find their way into the
workplace (Harris, Ives, & Junglas, 2012). Employees introduce services that they
use privately into their work environment and thereby undergo centrally provided
solutions. This development is supposed to drive innovation, productivity and
satisfaction (Harris et al., 2012). Thereby, business IT becomes more and more
similar to consumer IT and the lines between these two categories begin to blur.
Second, a current survey among 1,000 executives and IT decision makers revealed
that enterprise buyers are more and more mimicking consumers in their behavior
(avanade, 2013). Consequently, not only do the consumer services become more
important, the influence of individuals on the technological decisions increases due
to consumerization and changed buying processes. The market for the consumer
Authors / Viability of Cloud Services
6
cloud services in this study is therefore not limited to individual users but reaches far
into the enterprise sphere.
The expected contributions of this paper are threefold. First, our goal is to nudge
research away from focusing solely on one single KPI towards a theory that explains
the multidimensional nature of cloud service viability. Our results reveal complex
interrelationships between the different sub domains which need to be incorporated
when studying cloud service viability. Conversely, focusing on one performance
indicator can lead to decisions that harm other performance indicators such as
certain types of revenue streams. Second, our goal is to show that it is necessary to
differentiate between two different types of consumers’ willingness to pay. Although
paying for an existing service (that has been free before) is obviously largely
different to paying for a service upgrade, studies have equally referred to both types
as “willingness to pay”, without making an explicit statement what willingness to pay
refers to. Since our results highlight that these two revenue sources have different
antecedents, cloud services have to apply different strategies for improving these
KPIs. Third, our goal is to highlight the importance of relational factors for
understanding user behavior in the context of cloud services. As cloud service
relationships are characterized by information asymmetries, users’ uncertainty
perceptions have to be taken into account when studying user behavior in the cloud
context.
The remaining parts of the paper are structured as follows. In the next section, we
formulate the focus of our theoretical analysis. We then apply this perspective to
prior research on post-adoption and post-consumption phenomena to develop a
model explaining and predicting the viability of cloud services. Section three
Authors / Viability of Cloud Services
7
introduces our survey research methodology followed by a presentation of the
results in section four. Finally, section five discusses implications and avenues for
future research.
The Viability of Cloud Services
Cloud computing can be seen as evolution of IT service provisioning with respect to
both the underlying technology and the business models for delivering IT-based
solutions (Iyer & Henderson, 2010; Venters & Whitley, 2012). We define cloud
computing as a virtualization-based style of computing where IT resources are
offered in a highly scalable way as a cloud service over the internet (Armbrust et al.,
2010). Cloud services can be classified according to their deployment, service, and
revenue model.
We can distinguish two basic models of cloud service deployment, namely private
and public clouds. Private clouds are devoted to a single company only. They may
be built, owned, and managed by the organization or by a third party (Mell &
Grance, 2011). While they offer the highest degree of control over performance,
security, and reliability, they are often criticized for being similar to traditional
proprietary data centers without the typical advantages of clouds like no-up front
capital costs (Q. Zhang, Cheng, & Boutaba, 2010). Public cloud services are
available to the general public. They are owned, built, and managed by third parties.
While there is no fundamental difference in the technical realization compared to
private clouds, the consumer’s control over data, network and security is limited (Q.
Zhang et al., 2010). Within our study, we focus solely on public clouds built, owned,
and managed by an external cloud provider.
Authors / Viability of Cloud Services
8
There are essentially three service models for cloud-based solutions which offer
different levels of abstractions: infrastructure as a service (IaaS), platform as a
service (PaaS) and software as a service (SaaS) (Mell & Grance, 2011). IaaS refers
to the provision of hardware resources like processing, storage and networks. PaaS
solutions provide - in addition to the infrastructure level - a cloud software
development environment which can be used to develop SaaS solutions. SaaS
offers complete applications running on cloud infrastructure which is completely
managed and controlled by the provider (Benlian, Koufaris, & Hess, 2011). As they
are hosted on the internet, they are accessed through a web browser or a thin client
instead of being deployed on the user's computer. Within our study, we focus on
SaaS solutions.
The most common revenue model for cloud services is the freemium model that
implies offering basic functions for free but skimming off profit for advanced features.
Since many internet customers expect basic services to be free of charge, the
freemium model has gained enormous popularity and has also been adopted by a
variety of social networking or news platforms (Niculescu & Wu, 2014). Cloud
markets are characterized by the effect that the best providers capture a significantly
large share of the rewards with remaining competitors being left with little. In these
types of markets, cloud providers have to balance two goals, building and retaining
a large customer base and skimming customers’ willingness to pay. Viable cloud
services are successful in balancing these often divergent goals.
Authors / Viability of Cloud Services
9
Key Performance Indicators of Viable Cloud Services
Based on marketing and practitioner literature, we identified five consumer-related
KPIs – customer satisfaction, loyalty, word-of-mouth (WOM), willingness to pay for
retention (WTPR) and willingness to pay for an upgrade (WTPU) – that measure the
viability of a cloud service, i.e., its ability to build and retain a solid customer base
and transform users into paying customers. We excluded other customer-related
outcomes from literature such as repurchasing intentions or complaining behaviors
(Gustafsson & Johnson, 2004; M. D. Johnson, Anderson, & Fornell, 1995; Luo &
Homburg, 2007; Szymanski & Henard, 2001) since they are either not applicable to
cloud services or they have no influence on the services’ viability as defined above.
Therefore, we conducted an extensive, cross-disciplinary literature review (Webster
& Watson, 2002) to establish a detailed overview of the studied relationships,
contexts, examination objects and the domains of the previously specified
performance indicators (cf. Appendix A). We make use of these insights to inform
our hypothesis building and to discuss our findings in the light of previous research.
In the following, customer satisfaction, loyalty, WOM, WTPR and WTPU are clearly
defined and their commercial desirability is highlighted.
Customer satisfaction represents an important cornerstone for customer-oriented
businesses since it drives strategically important outcomes (Szymanski & Henard,
2001). In a recent survey, customers of cloud providers declare that contentment
with the services is the main reason why they have not changed their provider
suggesting that customer satisfaction is a key performance indicator for viable cloud
services (Redshift Research, 2012). Customer satisfaction furthermore is a core
construct in information systems (D. J. Kim, Ferrin, & Rao, 2009; S. S. Kim & Son,
Authors / Viability of Cloud Services
10
2009) and marketing research (Homburg, Koschate, & Hoyer, 2005; S. O. Olsen,
2002). With respect to scope and level of abstraction, two general types of customer
satisfaction are distinguished in the literature, namely transaction-specific customer
satisfaction and cumulative customer satisfaction (M. D. Johnson et al., 1995). While
transaction-specific satisfaction deals with the ex-post evaluation of a particular
product or service experience, cumulative customer satisfaction is a more abstract
construct that describes customers' total performance experience of a service
provider to date (Gelbrich & Roschk, 2011). Since we aim to study customer
satisfaction in the context of an ongoing cloud service experience, we adapt the
cumulative conceptualization in the following (R. L. Oliver, 1980). Accordingly, we
define satisfaction as customers’ subjective judgment resulting from positive and
negative observations of a cloud provider’s performance (R. L. Oliver, 1993).
Customer loyalty is a customer’s or user’s overall attachment or deep commitment
to a product, service, brand, or organization (R. L. Oliver, 1999). The concept is
described as a customer’s intention to continue using (continuance) a product in the
IT innovation literature (e.g., Cyr, 2008) or as repeated patronage in the marketing
literature (e.g., Lam, Shankar, Erramilli, & Murthy, 2004). Transferring this
conceptualization to the context of cloud services, we define loyalty as a customer’s
affective commitment to continue using the cloud service of a given provider.
Customer loyalty is an important indicator for the viability of cloud services because
it determines how well the current customer base can be retained. Also cloud
practice suggests that cloud providers need to become better at holding on to
customers since the “payoff takes longer—and because it is easier for customers to
switch providers” (Bain, 2012, p. 8).
Authors / Viability of Cloud Services
11
WOM is a “dominant force in the marketplace” (Mangold, Miller, & Brockway, 1999,
p. 73) and an “effective mean to increase the revenues and profits of firms” (S. S.
Kim & Son, 2009, p. 50). The growing presence of the internet is even expanding its
importance for the market success of IT services (Brown, Barry, Dacin, & Gunst,
2005). Compared to traditional software products, cloud services are often promoted
by a “word-of-mouth model” (Deloitte, 2009, p. 55). WOM refers to “informal
communication between private parties concerning evaluations of goods and
services” (Anderson, 1998, p. 6) which can be either positive, neutral or negative.
The additional benefit of an increasing customer base for the individual user resides
in improved opportunities of file sharing or – in some cases – the earning of more
storage. In line with previous research, we use positive WOM behavior – referring to
the customer intention to spread favorable information about the service provider
and its service among peers (Maxham III & Netemeyer, 2003) – as a proxy for
estimating the potential increase of the customer base. Regardless of the channel
through which WOM activities are distributed, we believe that any positive WOM
activity contributes to the viability of a cloud service because it influences how easy
and effective the network externalities inhibited in cloud services can be exploited by
the cloud provider.
Customer’s willingness to pay (WTP) is very valuable information necessary to
formulate a business strategy. Therefore, the challenge of its determination has long
been in focus of research and practice (Miller, Hofstetter, Krohmer, & Zhang, 2011).
For cloud providers, mostly using a freemium revenue model, this question is even
more important since they depend on customers who upgrade their service. In the
IS literature, WTP has been accordingly defined as a customer’s willingness to pay
Authors / Viability of Cloud Services
12
a small fee for advanced features of a service currently available for free (S. S. Kim
& Son, 2009). However, a second possibility to generate financial earnings is often
either ignored or used synonymously: the willingness to pay for retention, defined as
the willingness to pay for the same service currently available for free (Vock, van
Dolen, & de Ruyter, 2013). We differentiate these two types of WTP in our study.
The importance of WTP of any kind as an indicator for the long term viability of cloud
services is unquestioned. It determines how well current customers using the free
version can be converted into paying customers who actually generate revenues.
Drivers of Customer Satisfaction
Prior research suggests that customer satisfaction can be well explained by using
metrics derived from the technology acceptance model (Devaraj, Fan, & Kohli,
2002). While we expect these well tested relationships to also hold in our context,
we introduce perceived uncertainty as an important new driver of satisfaction for
cloud services. More particularly, we propose that low levels of uncertainty
perceptions are a premise for higher levels of customer satisfaction. The usage of
cloud services is accompanied by a loss of control by the user (Armbrust et al.,
2010). This loss of control can lead to perceived uncertainty. Uncertainty
perceptions are based on information asymmetries and might include users’ privacy,
security or availability concerns (Trenz, Huntgeburth, & Veit, 2013). In particular, it is
difficult for the consumer to judge whether his data is not misused, but stored
securely, and whether the necessary resource buffers are provided before capacity
overload incidents occur. Accordingly, we argue that the ongoing information
asymmetry in cloud user-provider relationships causes uncertainty to be also crucial
for customer satisfaction:
Authors / Viability of Cloud Services
13
H1: Consumers’ perceived uncertainty of using the service is negatively associated
with their level of satisfaction.
Drivers of Customer Loyalty
We find compelling evidence in the literature for the relationship between customer
satisfaction and loyalty (Cyr, 2008; D. J. Kim et al., 2009; J. Kim, Lee, Han, & Lee,
2002; Lam et al., 2004; Oliva, Oliver, & MacMillan, 1992; S. O. Olsen, 2002; Otim &
Grover, 2006). Other studies argue that although loyal customers are mostly
satisfied, the opposite does not have to be true (R. L. Oliver, 1999). However, such
inconsistencies can be largely explained by definitions and scopes of satisfaction
and loyalty that differ from our conceptualization (Han, Kwortnik, & Wang, 2008).
Consistent with the expectation-confirmation paradigm, we argue that satisfaction
with a cloud service is a key to building and retaining a loyal base of long-term
customers. In contrast, when customers become dissatisfied with the service they
will less likely continue using the service any longer (R. L. Oliver, 1980):
H2: Consumers’ level of satisfaction with the service is positively associated with
their loyalty.
Drivers of Word-of-Mouth
The link between customer satisfaction and WOM has been under investigation both
empirically and theoretically (Brady, Voorhees, & Brusco, 2012; Chiou, Droge, &
Hanvanich, 2002; Gittell, 2002; Heitmann, Lehmann, & Herrmann, 2007; Hennig-
Thurau, Gwinner, & Gremler, 2002; M. W. Johnson, Christensen, & Kagermann,
2008). A key motivation for WOM is a consumer’s experience with the services. This
service experience produces “a tension which is not eased by the use of the product
Authors / Viability of Cloud Services
14
alone, but must be channeled by way of talk, recommendation, and enthusiasm to
restore the balance” (Dichter, 1966, p. 148). Thus, affective states of either valence
stimulate WOM transmissions (Westbrook, 1987) and satisfied consumers are likely
to engage in positive WOM (Gittell, 2002). We believe that in the context of cloud
services this relationship also holds and thus, propose:
H3a: Consumers’ satisfaction with the service is positively associated with their level
of word-of-mouth.
Customer loyalty has not only been discussed as a customer-related outcome of
customer satisfaction but also as a driver of WOM. In context of online services, Kim
and Son (2009) provide evidence that a person’s dedication with the service is a
necessary precondition for positive WOM. They argue that since referring a peer
puts the customer socially at risk, positive WOM does not occur without a high level
of loyalty and dedication for the service provider. We believe that this relationship
also holds in our context. Moreover, cloud services exhibit strong network effects,
i.e., the value of the cloud service for customers depends on the number of others
using it (Katz & Shapiro, 1986). When users are intending to continue using the
service, they also have an incentive to increase the customer base through positive
WOM:
H3b: Consumers’ loyalty with the service is positively associated with their level of
word-of-mouth.
Drivers of Willingness to Pay
Compared to loyalty and WOM, the relationship between customer satisfaction and
WTP has attracted less attention in the literature despite its importance as a key
Authors / Viability of Cloud Services
15
element of the profit equation and its link to profitability (Homburg et al., 2005).
Based on equity theory (Adams, 1964), if unequal outcomes of a transaction occur
between customer and provider, individuals try to change certain parameters of the
exchange and try to establish a balance. Accordingly, a higher level of satisfaction
with the service implies a higher outcome for the customer which should also relate
to a higher level of outcome, in terms of payment, for the seller. Empirical support
for this theoretical argument is provided by Homburg et al. (2005) in the service
context. Such considerations are supposed to be even stronger when the customer
feels dedicated to a particular firm. If customers feel loyal to the cloud service, they
prefer to deal with this vendor as opposed to another service provider and
accordingly are willing to pay more (Palmatier, Scheer, & Steenkamp, 2007).
Previous studies have addressed different benefits for which customers could be
charged. Kim and Son (2009) address the willingness to pay for the same product or
service. Vock et al. (2013) investigate willingness to pay for advanced features or
additional purchases. Although paying for an existing service (that has been free
before) is obviously largely different to paying for a service upgrade, studies have
equally referred to both types as “willingness to pay”, without making an explicit
statement what willingness to pay refers to. Furthermore, no research study has
investigated these two types of willingness to pay simultaneously (cp. Appendix A).
In the context of cloud services, we argue that it is necessary to distinguish these
two types of willingness to pay carefully, because they depict two different paths to
financial success whose connection to the other dimensions of cloud viability is not
identical. In the following, we refer to willingness to pay for an upgrade (WTPU) if we
are discussing the willingness to pay for advanced features of a service such as
Authors / Viability of Cloud Services
16
more storage or advanced security. Accordingly, we refer to willingness to pay for
retention (WTPR) if we are discussing consumers’ willingness to pay for keeping the
same service level that has been free before.
While both a high level of customer satisfaction and loyalty have been found to
positively influence classical WTP in previous research (Fullerton, 2003; Homburg et
al., 2005), we propose that in the cloud service context customer satisfaction does
not drive consumers to pay for advanced features (WTPU). The major driver of this
difference is the exploitation of the freemium model as explained in the following.
Basic functions are usually offered for free. Generally, consumers are willing to pay
more if they expect a higher utility from the transaction. However, if consumers are
highly satisfied with the current service level, the urge to change their present
service configuration is low. Accordingly, they derive less additional value from an
extended service than customers whose needs are currently not fully satisfied. This
difference in the valuation of a service upgrade is translated into differences in the
willingness to pay for advanced features. Customers who are satisfied tend to stay
with the present service configuration and feel no necessity to pay a fee for
advanced features. In contrast, customers with lower levels of satisfaction with the
basic service possess a comparably higher need for additional features and
therefore have a higher WPTU. Therefore, we propose:
H4a: Consumers’ satisfaction with the service is negatively associated with their
WTPU.
The opposite mechanism applies to the case of WTPR. WTPR is important when
the company decides to stop offering the current service level for free. Consumers
who are highly satisfied with the current service will perceive a higher financial or
Authors / Viability of Cloud Services
17
psychological loss in utility. This opportunity cost can be translated into a WTP for
retaining the current state. In contrast, it is easier for users with less favorable
feelings toward the service to leave the provider. Accordingly, their willingness to
procure money to preserve the current service level is lower. This argumentation is
in line with Homburg et al. (2005) who argue that consumers search a balance
between the outcome of an transaction (satisfaction) and the input of an transaction
(payment). Therefore, satisfied consumers should be willing to pay more for the
retention of the service than unsatisfied consumers:
H5a: Consumers’ satisfaction with the service is positively associated with their
WTPR.
A strong dedication for the service influences how consumers react when they have
the opportunity to intensify their relationship. Their dedication for the service makes
loyal customers more open for upgrading their service configuration even if they are
charged an additional fee by the provider (WTPU). Accordingly, they will refer to the
provider if they need additional features. Loyalty decreases consumers’ price
sensitivity (Krishnamurthi & Raj, 1991) and therefore increases the willingness to
intensify the relationship with the provider at a cost. Therefore, we assume that loyal
customers have a higher WTPU than consumers without a dedication for the cloud
service:
H4b: Consumers’ loyalty with the service is positively associated with their WTPU.
Loyal consumers are also willing to pay more for the retention of the service than
uncommitted consumers. Due to their high dedication with the service, they perceive
high costs to terminate this strong relationship with the service that can be described
Authors / Viability of Cloud Services
18
as types of switching costs (S. S. Kim & Son, 2009) or search costs (Reichheld &
Sasser, 1990). Therefore, they are willing to invest money to inhibit the occurrence
of these switching costs. However, consumers who are less committed do not
perceive this harm and therefore are willing to pay less for the retention of the
service:
H5b: Consumers’ loyalty with the service is positively associated with their WTPR.
Apart from these relationships with the other KPIs, the most common view to
establish willingness to pay is to focus on the utility that the customer derives from
the service (Miller et al., 2011). For the free service this utility directly derives from
the usefulness of the currently experienced free service. Accordingly, a higher
perceived usefulness of a particular service should lead to a higher WTPR:
H5c: Consumers’ perceived usefulness of the service is positively associated with
their WTPR.
In turn, for the premium service, this utility derives from the usefulness of the basic
service plus the perceived value of the premium services (Dodds, Monroe, &
Grewal, 1991). Accordingly, both perceived usefulness and perceived value of
upgrade are driving consumers WTPU:
H4c: Consumers’ perceived usefulness of the service is positively associated with
their WTPU.
H4d: Consumers’ perceived value of the service upgrade is positively associated
with their WTPU.
Figure 1 presents an overview of our research model.
Authors / Viability of Cloud Services
19
Research Methodology
The hypotheses derived in the previous section were tested using survey data from
an online questionnaire among actual users of cloud storage services. We focus on
cloud storage services because these services are widely adopted by internet users
(Zetta, 2010) and share the typical characteristics of other cloud-based services
(e.g., appearance of infinite computing resources available on demand, elimination
of an up-front commitment, ability to pay for use of computing resources, see
Armbrust et al., 2010). Moreover, they are characterized by very low marginal costs
Figure 1. Research Model and Proposed Hypotheses
Authors / Viability of Cloud Services
20
and therefore have to address the tension between growth and revenue generation.
Their cost structure and the highly competitive situation in growing cloud markets
provide incentives to offer basic functions like file-sharing, synchronization and a
certain amount of storage for free. However, cloud storage services also need to
identify ways to generate revenues. In the following, we describe our measurement
development as well as the survey deployment and data collection procedures.
Measurement Development
All measures used in our study were adopted from existing measures. However,
they were adapted to the context of our study. On grounds of the critique raised
about the validation of scales in the IS discipline (e.g., Boudreau, Gefen, & Straub,
2001; Scott B. MacKenzie, Podsakoff, & Podsakoff, 2011), we decided to re-validate
our constructs. This process included the definition and assessment of the domain
and dimensionality of the constructs using two sorting procedures (Moore &
Benbasat, 1991) and the assessment of content validity using a rating method
(Hinkin & Tracey, 1999; Scott B. MacKenzie et al., 2011). We pilot tested the
preliminary instrument 196 participants. After the pre-test, the respondents were
asked to give open feedback regarding composition of the survey, overall time, and
other issues they experienced. Following the pre-test, the instrument was shortened,
refined, and validated for its statistical properties. In the final survey, all principal
constructs were measured as first-order reflective constructs using three or more
indicators. An overview of all measures and their sources is given in appendix B.
Authors / Viability of Cloud Services
21
Survey Deployment and Data Collection
We collected our data using an online survey, since the regular online access is a
prerequisite for usage of such a service. While little is known about which part of the
internet population is using cloud storage services, a representative set (with
respect to gender and age) of internet users that matches the general population in
Germany was pre-selected (cf. AGOF 2013) and subsequently, only those
participants of the survey were surveyed that use the market-leading cloud storage
service Dropbox - ensuring comparability of responses. Using these requirements,
a professional online panel has sent individual invitations to its members in the
period between 12th of November and 9th of December 2012. Overall, we received
2.011 valid responses of which 638 declared to use Dropbox. We further eliminated
responses of those Dropbox users that declared to use the premium service (42
premium users) - again to ensure comparability of responses. By the end, 596
responses were deemed useful for the subsequent analysis. Table 1 summarizes
the structure of the sample.
Authors / Viability of Cloud Services
22
Data Analysis and Results
We used partial least squares structural equation modelling (PLS-SEM) to validate
the structural model and test our hypotheses. Two different types of SEM
approaches exist, covariance-based SEM (CB-SEM) and partial least square SEM
(PLS-SEM), which differ in their underlying philosophy and estimation objectives
(Gefen, Rigdon, & Straub, 2011). On the one hand, CB-SEM emphasizes how well
the proposed research model accounts for measurement item co-variances, thereby
offering various indices how well parameter estimates match sample co-variances
Table 1. Years of Age and Gender Distribution among Respondents
Years of age/gender Representative sample of internet users
Users of free Dropbox version (our final sample)
Number Quota Number Quota
14-19/female 90 5% 41 7%
20-29/female 186 9% 105 18%
30-39/female 170 9% 33 6%
40-49/female 210 10% 24 4%
50-59/female 166 8% 19 3%
60-69/female 132 7% 14 2%
14-19/male 95 5% 53 9%
20-29/male 189 9% 113 19%
30-39/male 181 9% 75 13%
40-49/male 233 12% 54 9%
50-59/male 180 9% 36 6%
60-69/male 179 9% 29 5%
Overall 2,011 596
Authors / Viability of Cloud Services
23
(Chin, 1998). On the other hand, PLS-SEM uses the empirical data for estimating
relationships with the aim to maximize the explained variance in the endogenous
latent variable (Hair, Hult, Ringle, & Sarstedt, 2014). The primarily goal is therefore
to predict and explain variance.
There is an ongoing controversial debate about which SEM tool to use as well as
CB-SEM’s and PLS-SEM’s relative ability to support the empirical evaluation of
hypothesized relationships among variables (Goodhue, Lewis, & Thompson, 2006,
2012; Marcoulides, Chin, & Saunders, 2012). Two popular reasons for choosing
PLS-SEM were that it provides more accurate results for studies with small sample
sizes and non-normally distributed variables (Ringle, Sarstedt, & Straub, 2012).
However, this assertion has not been confirmed by previous comparison studies.
Rather, a recent Monte Carlo simulation-based study of Goodhue et al. (2012)
shows that CB-SEM and PLS-SEM provide consistent results regarding testing
relationships between variables:
“[…] if one is in the early stages of a research investigation and is
concerned more with identifying potential relationships than the
magnitude of those relationships, then regression or PLS would be
appropriate. As the research stream progresses and accuracy of the
estimates becomes more important, LISREL (or other CB-SEM
techniques) would likely be preferable” (Goodhue et al., 2012, p. 999).
To sum the discourse up, the choice of the SEM should primarily depend on the
research objective. Thereby, PLS-SEM is more suitable for exploratory and CB-SEM
is more suitable for explanatory evaluations of theoretical systems. Given the early
stage of this investigation, the exploratory character of the study and the primary
Authors / Viability of Cloud Services
24
interest in identifying potential relationships between variables, we decided to use
PLS-SEM for evaluating the drivers and interrelationships of our KPIs. However,
consistent empirical results are expected when using CB-SEM.
Descriptive Statistics of Sample
Table 2 depicts the descriptive statistics of the surveyed Dropbox users. The
statistics highlight that the sample consists of heterogeneous sub-groups of low and
highly educated, employed and unemployed, low and high income as well as male
and female respondents. This provides evidence that no demographic group was
systematically excluded from the study.
Table 2. Descriptive Statistics of Dropbox Users (free version)
Education
No education
Secondary school
Higher education
Completed vocational
training
University degree
Doctorate degree
2 (0.3%) 129 (21.6%) 192 (30.0%) 108 (18.1%) 183 (30.7%) 4 (0.7%)
Income
< €500 €501-€1,500 €1,501-€2,500 €2,501-€3,500 > €3,500 Not specified
51 (8.6%) 121 (20.3%) 147 (24.7%) 92 (15.4%) 86 (14.4%) 94 (15.8%)
Occupation
In training
Employed Unemployed
or retired Not specified
214 (36.4%)
307 (51.5%) 72 (12.1%) 3 (0.5%)
Measurement Validation
For reflective measurement models in PLS-SEM, there are three criteria for
evaluating the reliability and validity of the measurement models, namely internal
consistency reliability, convergent validity and discriminant validity (Henseler,
Ringle, & Sinkovics, 2009). We checked internal consistency using composite
reliability scores (Table 3). All measurement models exhibit satisfactory composite
Authors / Viability of Cloud Services
25
reliability values above 0.7 (Nunnally & Bernstein, 1994). At the construct level, we
assessed convergent validity based on indicator reliability and the average variance
extracted (AVE) (Henseler et al., 2009). All outer loadings were above the
recommended threshold of 0.708 suggesting reliable indicators. Moreover, all
constructs had AVE values above 0.5 suggesting that more than half of the item’s
variance is explained by the latent variable. We evaluated discriminant validity
based on the Fornell-Larcker criterion (Fornell & Larcker, 1981) as well as
comparing outer and cross loadings (Hair et al. 2011). According to Fornell-Larcker
criterion, the square root of each latent variable’s AVE must be larger than its
correlation with any other latent variable. This is the case for our measurement
models (cf. appendix C). Moreover, for each item the outer loading on its associated
latent variable was higher than the cross loadings on all other latent variables (cf.
appendix D). Since both criteria are fulfilled, one can conclude that the
measurement models exhibit discriminant validity.
Table 3. Measurement Model Results
Con-structs
Variable Name
Outer Loading
Items per Construct
AVE Composite Reliability
Mean Standard Deviation
Uncertainty UNC1 UNC2 UNC3 UNC4
0.8695 0.9005 0.9495 0.9413
4 0.84 0.95 3.48 1.51
Ease of Use PEU1
PEU2
PEU3
PEU4
0.9227 0.7195 0.8166 0.9162
4 0.72 0.91 5.48 1.16
Usefulness PU1 PU2 PU3 PU4
0.9289 0.9335 0.9133 0.9006
4 0.84 0.96 4.75 1.51
Value of Upgrade
PVU1 PVU2 PVU3
0.9584 0.9551 0.9584
3 0.92 0.97 2.67 1.85
Satisfaction SAT1 SAT2 SAT3 SAT4
0.9146 0.8953 0.9286 0.8492
4 0.81 0.94 5.76 1.11
Authors / Viability of Cloud Services
26
WOM WOM1 WOM2 WOM3 WOM4
0.9183 0.8949 0.893 0.853
4 0.79 0.94 5.00 1.56
Loyalty LOY1 LOY2 LOY3 LOY4
0.8584 0.8552 0.9015 0.9285
4 0.79 0.94 5.24 1.43
WTPR WTPR1 WTPR2 WTPR4 WTPR4
0.686 (*) 0.8313 0.8944 0.8886
3 0.76 0.90 14.35 26.25
WTPU WTPU1 WTPU2 WTPU3 WTPU4
0.9216 0.9203 0.8272 0.8878
4 0.79 0.94 4.14 12.33
IT Experience
ITE1 ITE2 ITE3 ITE4
0.8678 0.9107 0.8851 0.8543
4 0.77 0.93 5.33 1.13
Internet Use IUSE - 1 - - 5.00 3.23
Cloud Knowledge
CK1 CK2 CK3 CK4
0.932 0.9458 0.9261 0.9141
4 0.86 0.96 4.27 1.46
Age AGE - 1 - - 33.19 13.88
Gender GEN - 1 - - 0.60 0.49
Income INC - 1 - - 4.61 2.10
Because data for each respondent was obtained using a single measurement
method, we applied the recommended procedural and statistical remedies as
proposed by Podsakoff et al. (2003) to minimize and control for common method
variance (CMV). First, we conducted the Harman’s single factor test using an
exploratory factor analysis in SPSS (Podsakoff et al., 2003). The unrotated principal
component factor analysis revealed eight factors with eigenvalues above 1,
explaining 80% of the variance. The most prominent component accounted for 34%
of the variance. Since neither a single factor emerged, nor one general factor
accounts for the majority of the covariance among the variables, evidence is
provided that CMV did not bias the results (Malhotra, Kim, & Patil, 2006). Second,
we used the marker-variable technique as proposed by Lindell and Whitney (2001)
Authors / Viability of Cloud Services
27
to examine how a potential CMV biases the results. The marker-variable technique
controls for CMV by including “a measure of the assumed source of method
variance as a covariate in the statistical analysis” (Podsakoff et al., 2003, p. 889).
The technique was used in a post-hoc manner by taking the second smallest
correlation among the variables to be the extent of CMV (0.027 between UNC
WOM, cf. Appendix C). We calculated the CMV-adjusted estimate and the test
statistic for each pair of the correlation matrix (using equations (4) and (5) in Lindell
& Whitney, 2001, p. 116). Because the CMV-estimate (second smallest correlation)
was close to zero, almost all previously significant correlations remained significant
(only UNCWOM turned insignificant). Thus, the marker-variable technique also
suggests that CMV did not bias the results. Third, we reconstructed the latent
variables and measures in AMOS 22 to include a latent general common method
factor that was allowed to load on every item in the research model (Podsakoff et
al., 2003). Since the common variance between the measures and the latent
general common method factor was also close to zero, we conclude that the path
estimates and significance levels are neither substantially inflated nor deflated by
CMV (S. B. MacKenzie, Podsakoff, & Paine, 1999). Overall, based on these three
statistical tests, we can rule out concerns that CMV biases the results.
Testing the Structural Model
Once evidence is provided that the measurement instrument is both reliable and
valid, the next step is to evaluate the path coefficients and their significance in the
structural model (Hair et al., 2014). As the PLS algorithm assumes that there is no
collinearity between the exogenous latent variables, one needs to ensure that this is
the case for each endogenous latent variable. Using the latent variables scores of
Authors / Viability of Cloud Services
28
each variable, we ran a linear regression using SPSS and calculated the variance
inflation factor (VIF) based on the regression (Mooi & Sarstedt, 2011). The VIF
measures how much of the variance of an estimated regression coefficient is
increased because of collinearity, i.e., because two exogenous latent variables are
correlated. The VIF for all correlations was far below the recommended threshold of
5 (even all VIF were below 2.5) suggesting that multicollinearity did not bias path
coefficient estimations.
We included control variables into our structural model. Beyond age, gender and
income as demographic variables, we also included IT experience, cloud knowledge
and internet use as additional covariates into our model to check whether the effects
can be explained by differences in the users’ level of experience with technology,
computer or the internet. Based on 5.000 bootstrap samples, we found significant
effects of age (b=.082; p<.01) and gender (b=-.064 p<.05) on loyalty. Furthermore,
IT experience and cloud knowledge both had a strong effect on satisfaction. While
users with high knowledge about cloud tend to possess lower levels of satisfaction
(b=-.109; p<0.01), users with high experience with using IT are more satisfied with
the service (b=.133; p<.01). Also women were more satisfied with the service but
willing to pay less for retention than men (b=-.118; p<0.01; b=.136; p<0.01). Finally,
income had a significant effect on WOM (b=-.062; p<0.05) as well as on WTPU
(b=.108; p<0.01). Despite several significant effects, the control variables had no
effect on the conclusions drawn from the structural model evaluation, i.e., all
hypothesized effects remained within the same level and their significance was not
influenced by the control variables.
Authors / Viability of Cloud Services
29
Regarding our hypotheses, we found that the negative relationship between
uncertainty and satisfaction was highly significant (b=-.16; p<.001; H1), over and
above the well-established constructs of perceived usefulness (b=.28; p<.001) and
perceived ease of use (b=.47, p<.001). Accordingly, all three exogenous variables
had strong loadings on satisfaction and explained 56% of its variance. Furthermore,
as hypothesized, satisfaction had a significant positive effect on loyalty with the
cloud service (b=.58; p<.001; H2). The variance explained for loyalty was 38%.
Regarding WOM, the relationship with satisfaction (b=.31; p<.001; H3a) was
confirmed as well as for H3b, claiming that loyalty increases word-of-mouth (b=.38;
p<.001; H3b). 41% of the variance of WOM was explained by our model. Regarding
WTP, our study shows that the connection of WTPR and WTPU to the other
dimensions of cloud viability is not identical. On the one hand, WTPU is significantly
negatively influenced by satisfaction (b=-.10; p<.01; H4a) and positively affected by
loyalty (b=.10; p<.01; H4b), perceived usefulness (b=.08; p<.01; H4c) and perceived
value of upgrade (b=.41; p<.001; H4d), explaining 23% of the variance. On the other
hand, WTPR is significantly influenced by loyalty (b=0.14, p<.01; H5b) and
perceived usefulness (b=.16; p<.001; H5c), while satisfaction has no significant
positive impact (b=.04; p=.271; H5a). Therefore, H5a was not supported in our
study. Figure 2 depicts the overall results of the structural model test.
Authors / Viability of Cloud Services
30
Figure 2. Structural Model Results
Discussion
The objective of this study is to develop and test a parsimonious model that
examines the drivers of five KPIs of viable cloud services. In the emerging context of
cloud services, we combine established and new, cloud-specific drivers of each
performance indicator and investigate their influence on each other. Thereby, we
were able to resolve inconsistencies among previous studies regarding the
relationships between satisfaction, loyalty, WOM, WTPR and WTPU in this new
theoretical context. Overall, our findings provide strong support for our research
model. The three major findings that provide new insights on viable cloud services
Authors / Viability of Cloud Services
31
are presented in detail in the following. Subsequently, the theoretical and practical
contributions as well as limitations of our study and future research are discussed.
Key Findings
The cloud-specific driver, perceived uncertainty is well-suited to explain customer
satisfaction in the context of cloud services, even after controlling for the well-
established drivers of perceived usefulness and perceived ease of use. This is
consistent with the expectation-confirmation paradigm where satisfaction is said to
be formed based on the expectations and experiences with the services
(Bhattacherjee, 2001; Hong, Thong, & Tam, 2006). Hereby, users’ believes about
the service (i.e., perceived uncertainty, usefulness and ease of use) are a function of
the expectations and subsequent experiences with the service (R. Oliver, 1977).
This study is the first to examine uncertainty perceptions as a major driver of
customer satisfaction in the context of cloud computing.
We find empirical support for loyalty as the strongest driver of WOM. This finding is
in line with the assumption that users are only willing to take the social risks of
recommending the cloud service when they are highly dedicated to the service as
highlighted by Kim & Son (2009) in their study of online services. However, we also
find strong empirical support for the positive relationship between satisfaction and
positive WOM (Brady et al., 2012; Heitmann et al., 2007; J. Zhang & Bloemer,
2008). This shows that satisfied customers have a tendency to share their positive
service experience with their peers (Arndt, 1967; Dichter, 1966). A possible
explanation for this strong relation between satisfaction and WOM is the nature of
the channel through which recommendations are distributed. The offline channel
Authors / Viability of Cloud Services
32
usually provides a wealth of social bonding or personal fortitude among sender and
receiver. These opportunities are absent in the online channel through which most
cloud service referrals are distributed (Dellarocas, 2003). Here, WOM spreads much
faster, is less personal and thus, is putting the customer’s social image less at risk
than in offline scenarios (Reichheld, 2003). Moreover, the additional benefit of an
increasing customer base (improved opportunities for file sharing and – in some
cases – more storage as an incentive) which motivates WOM activities is not limited
to loyal customers but is instead a goal of all users positively experiencing the
service. Thus, both satisfaction and loyalty drive WOM for cloud services.
Our last set of performance indicators, WTPR and WTPU, are extremely important
for providers in the context of cloud services as revenues are generated based on a
freemium revenue model (Teece, 2010). The deviation from treating WTP as a one-
dimensional concept enables us to develop more fine grained insights on the
potentials for revenue stream generation. Unlike previous marketing research
(Homburg et al., 2005), our study shows that customer satisfaction has only an
indirect effect on customers’ WTP for retention in the context of cloud services. Few
previous studies also found no support for the direct positive relationship between
customer satisfaction and WTP, e.g., in the contexts of consumer goods (J. Zhang &
Bloemer, 2008) and travel services (Homburg, Wieseke, & Hoyer, 2009). However,
these contexts are hardly comparable to our study. Moreover, we find a significant
negative relationship between satisfaction and willingness to pay for upgrade. This
finding implies that a high level of satisfaction can even have negative
consequences for the firm’s revenue, especially in a freemium environment, since
consumers that are very satisfied with their current service level have little incentive
Authors / Viability of Cloud Services
33
to invest financial means in additional features or capacity. The fact that we find no
effect of satisfaction on the second type of willingness to pay, WTPR, highlights the
importance to differentiate between the two. Overall, our findings indicate that for
cloud services, constructs other than customer satisfaction are needed to explain
revenue streams of cloud services. While previous research has mainly
concentrated on customer satisfaction as the central concept for increasing revenue
streams, our study reveals that loyalty is a key for cloud providers to yield profits.
What needs to be kept in mind: reaching customer loyalty is especially difficult for
cloud providers because they are hardly able to establish social bonding or personal
fortitude as common in offline service scenarios (R. Oliver, 1977). Therefore, when
loyalty becomes a key driver of such important business-outcomes like WTP, cloud
providers have to find alternative measures to generate revenue streams. One
starting point for providers is the customers’ usefulness perception of the service or
the upgrade option which is found to directly influence WTP consistent with previous
research.
Theoretical Contribution
Overall, our study makes three major contributions. First, our study shows that in the
context of cloud service, it is not sufficient to focus solely on one KPI. The complex
interrelationships between satisfaction, loyalty, WOM, WTPR and WTPU create the
necessity to move away from simple models focusing on single outcome variables
(cf. Appendix A). For instance, a focus on satisfaction as the main KPI would neglect
its divergent influence on customers’ WTP, which is an influential factor for viable
cloud services. This simplification would involve the danger of incorrect inferences
or policies. Our study implies that we need to develop theories that account for the
Authors / Viability of Cloud Services
34
multidimensional nature of cloud service viability and incorporate the
interrelationships between the different dimensions. Moreover, the mediating effects
identified in our study (SATLOYWTPR) should encourage research to re-
examine our five KPIs together in other theoretical contexts in order to uncover
spurious relationships wrongly inferred in previous studies.
Second, we introduce a precise conceptual differentiation between willingness to
pay for retention and willingness to pay for an upgrade. We provide empirical
evidence for their different interrelations with other KPIs. While previous studies
have used both types of willingness to pay synonymously, the differentiation
between the two strategies to generate revenue streams for cloud services is
important because they relate to different strategies that cloud providers can
employ, i.e., charging existing customers for their current service level or generating
additional needs via the free services that customers are willing to pay for. By
showing that the two types of willingness to pay have different antecedents, our
results imply that their conceptual and empirical separation is crucial to understand
different paths to cloud viability.
Third, our study highlights the importance of incorporating relational factors for
understanding user behavior and users’ service experience in the context of cloud
services. Previous research on technology adoption or continuance has mainly
looked at the perceived characteristics of the IT artifact (e.g., performance
expectancy, effort expectancy) or the social environment of the user (e.g., social
influence, facilitating conditions). When using cloud services however, customers
are dependent on the provider over the whole life-cycle of the business relationship.
As their relationship is continuously characterized by information asymmetries, the
Authors / Viability of Cloud Services
35
corresponding uncertainty has to be taken into account when studying customer
behavior in the cloud context. Our study shows that the characteristics of the cloud
provider-user relationship are increasingly shaping users’ experiences with the
service. These differ sharply from other service contexts, where the service
provisioning takes place within a limited time frame. Accordingly, our study
establishes users’ uncertainty perception as an important antecedent for explaining
user behavior in the context of cloud computing. We encourage future research to
be more strongly attentive to the characteristics of the cloud provider-user
relationship and study post-adoption phenomena using relational factors such as
users’ uncertainty perception.
Limitations and Suggestions for Future Research
First, cloud storage services were used as a study context for the evaluation.
Although cloud storage services are widely adopted by internet users and exhibit the
typical characteristics of cloud services, future research should re-examine viability
for other cloud services. Second, our study identified uncertainty as an important
inhibitor of satisfaction. Unfortunately the scope of our study did not allow us to
explain how the uncertainty connected to cloud services arises and how it could be
mitigated. However, this question should be addressed in further research because
it is of high theoretical and practical interest. Third, our findings regarding the effects
of satisfaction and loyalty on willingness to pay contradict previous studies. We
explain these findings by the unique characteristics of cloud services compared to
other contexts. However, this finding calls for further research challenging these
relationships in other scenarios and identifying contingency factors in order to create
Authors / Viability of Cloud Services
36
a broader understanding of the development of willingness to pay in different online
service scenarios.
Implications for Practice
Based on our results, we derive recommendations for three generic viability
strategies that can be applied in practice: development, retention and habituation
(see Table 4).
Table 4. Strategic Implications for Cloud Service Providers
Viability Strategy
Recommendations
Development Effective versioning of services with a clearly identifiable added value of premium service
Make sure that free version is useful but does not fully address all needs of the user
Build a loyal customer base which is strongly committed to the service
Retention Make sure that free version is highly useful for users
Only switch to subscription model if you have a large number of loyal customers
Habituation Effective versioning of services with a clearly identifiable added value of premium service
Offer long-term free trials of premium service to get user accustomed to premium service
End trial period if individual user is committed to service
Cloud providers pursuing the development viability strategy mainly aim for
generating revenue streams based on transforming free users into paying
customers and for extending the user base through free service offerings. The goal
of these services is to develop free users to become premium users. Based on our
results, providers using the development viability strategy have to design the free
and premium versions of the service in a way that the premium service is clearly
distinguishable from the free version and whose identifiable added value is desirable
to a broad audience. Preferably, some advanced user objectives cannot be
Authors / Viability of Cloud Services
37
achieved with the free version. This is for instance hardly the case for premium
services that offer an add-free interface. Beyond good versioning, providers have to
focus on building a loyal customer base with users that are unlikely to switch the
provider. Since there are typically very little points of contact with the customers,
these have to be exploited effectively (e.g., problem handling or requests using
social web technology) to tie users to the service provider. A good example of a
cloud service pursuing this “development” strategy is Prezi, a cloud presentation
software service for presenting ideas on a virtual canvas. The free version allows
users to create presentations that are publicly visible. Moreover, users are able to
collaborate and present on Prezi using a small amount of free cloud storage. While
the free version of Prezi is a useful tool for consumers, premium features like
privacy, editing presentations offline or more storage provide a clearly identifiable
added value for any customer. Moreover, the service has implemented various ways
to build a solid community of users around its application. Therefore, they are also
successful in building a loyal customer base.
Cloud providers pursuing the retention viability strategy mainly aim for generating
revenue streams based on switching to a subscription revenue model at an
opportune point of time. They use the free version to grow fast and monetize later.
Based on our results, providers trying to become viable using this strategy need to
make sure that the free version strongly increases the (work) performance of the
user. Moreover, providers should wait with switching to a subscription model until a
large number of users is affectively committed or faces high switching costs. A good
example of a viable cloud service that has switched to a subscription model after
having initially operated a freemium model is Chargify LCC as described at the
Authors / Viability of Cloud Services
38
beginning of this article. Chargify LCC was successful in this transformation
because they had a considerably large number of customers who felt affectively
committed to the service and perceived a clear usefulness of the service Chargify
offered. Therefore, Chargify successfully turned into a viable cloud service.
Cloud providers pursuing the habituation viability strategy aim for generating
revenues by skimming both types of willingness to pay. Thereby, they offer each
individual user a very long, possibly hidden, trial phase which offers certain premium
features for free. At an opportune time, they end the trial phase and bet on
customers who adjusted their preferences or their habits towards the premium
features and are therefore willing to pay for keeping the same service level that has
been free before. At the same time, they keep effective versions of the free and
premium service and try to develop free customers to become premium users. Apart
from the guidelines for the other two strategies that the habituation strategy borrows
from, our results suggest that the length of the trial period should be extensive,
allowing users to become strongly accustomed to the service. A good example of a
cloud service pursuing the hybrid viability strategy is Dropbox. Dropbox offers a free
account with a set storage size and paid subscriptions for accounts with more
capacity. In 2012, Dropbox has launched the program “The Great Space Race” that
let college students gaining up to 25GB of free storage space for two years. The
program was meant to increase Dropbox’s market share among students but at the
same time intended to accustom these users to using more storage than the free
version offers. During that time, users might have unconsciously changed their
behavior in using Dropbox towards a higher level of (storage) requirements, e.g., by
synchronizing more files that they otherwise would or changing their sharing
Authors / Viability of Cloud Services
39
behavior in collaborations. After the long period of two years, Dropbox can hope that
many trial users are willing to pay for keeping the same amount of storage capacity,
since their habits have changed and they do not want to remove files that they have
been able to access from everywhere or to end active collaborations with partners.
The habituation viability strategy therefore tries to combine the strength of the
retention and upgrade strategy.
The choice for a specific strategy depends on a market specific assessment whether
the drivers of our five KPIs can be successfully influenced or not. In any case, our
results provide specific recommendations that have been carved out through our
multidimensional conceptualization of cloud viability. These recommendations can
be used by cloud providers to develop a viable position in their particular competitive
cloud markets.
Authors / Viability of Cloud Services
40
References
Adams, J. S. (1964). Inequality in Social Exchange. In L. Berkowitz (Ed.), Advances
in Experimental Social Psychology (pp. 267–299). New York, NY: Academic
Press.
Agustin, C., & Singh, J. (2005). Curvilinear effects of consumer loyalty determinants
in relational exchanges. Journal of Marketing Research, 42(1), 96–108.
Anderson, E. W. (1998). Customer satisfaction and word of mouth. Journal of
Service Research, 1(1), 5–17.
Andreassen, T. W. (1999). What drives customer loyalty with complaint resolution?
Journal of Service Research, 1(4), 324 –332.
Arens, Z., & Rust, R. (2012). The duality of decisions and the case for impulsiveness
metrics. Journal of the Academy of Marketing Science, 40(3), 468–479.
Armbrust, M., Stoica, I., Zaharia, M., Fox, A., Griffith, R., Joseph, A. D., … Rabkin,
A. (2010). A view of cloud computing. Communications of the ACM, 53(4),
50–58.
Arndt, J. (1967). Role of product-related conversations in the diffusion of a new
product. Journal of Marketing Research, 4(3), 291–295.
avanade (2013). Global Survey: B2B is the New B2C. Retrieved June 1, 2014, from
http://www.avanade.com/us/approach/research/Pages/Global-survey-B2B-is-
the-new-B2C.aspx
Bain (2012). Selling the cloud. Retrieved January 24, 2013, from
http://www.bain.com/publications/articles/selling-the-cloud.aspx
Baskerville, R. (2011). Individual information systems as a research arena.
European Journal of Information Systems, 20(3), 251–254.
Authors / Viability of Cloud Services
41
Benlian, A., Koufaris, M., & Hess, T. (2011). Service quality in software-as-a-service:
developing the saas-qual measure and examining its role in usage
continuance. Journal of Management Information Systems, 28(3), 85–126.
Bhattacherjee, A. (2001). Understanding information systems continuance: an
expectation-confirmation model. MIS Quarterly, 25(3), 351–370.
Blocker, C., Flint, D., Myers, M., & Slater, S. (2011). Proactive customer orientation
and its role for creating customer value in global markets. Journal of the
Academy of Marketing Science, 39(2), 216–233.
Boudreau, M.-C., Gefen, D., & Straub, D. W. (2001). Validation in Information
Systems Research: A State-of-the-Art Assessment. MIS Quarterly, 25(1), 1–
16.
Brady, M. K., Voorhees, C. M., & Brusco, M. J. (2012). Service sweethearting: its
antecedents and customer consequences. Journal of Marketing, 76(2), 81–
98.
Brakus, J. J., Schmitt, B. H., & Zarantonello, L. (2009). Brand experience: what is it?
How is it measured? Does it affect loyalty? Journal of Marketing, 73(3), 52–
68.
Brown, T. J., Barry, T. E., Dacin, P. A., & Gunst, R. F. (2005). Spreading the word:
investigating antecedents of consumers’ positive word-of-mouth intentions
and behaviors in a retailing context. Journal of the Academy of Marketing
Science, 33(2), 123–138.
Chandrashekaran, M., Rotte, K., Tax, S. S., & Grewal, R. (2007). Satisfaction
strength and customer loyalty. Journal of Marketing Research, 44(1), 153–
163.
Authors / Viability of Cloud Services
42
Chestnut, B. (2010). Going Freemium: One year later. Retrieved May 15, 2014, from
http://blog.mailchimp.com/going-freemium-one-year-later/
Chin, W. W. (1998). Commentary: Issues and opinion on structural equation
Modeling. MIS Quarterly, 22(1), vii–xvi.
Chiou, J.-S., & Droge, C. (2006). Service quality, trust, specific asset investment,
and expertise: direct and indirect effects in a satisfaction-loyalty framework.
Journal of the Academy of Marketing Science, 34(4), 613 –627.
Chiou, J.-S., Droge, C., & Hanvanich, S. (2002). Does customer knowledge affect
how loyalty is formed? Journal of Service Research, 5(2), 113–124.
Chitturi, R., Raghunathan, R., & Mahajan, V. (2008). Delight by design: the role of
hedonic versus utilitarian benefits. Journal of Marketing, 72(3), 48–63.
Colgate, M. R., & Danaher, P. J. (2000). Implementing a customer relationship
strategy: the asymmetric impact of poor versus excellent execution. Journal
of the Academy of Marketing Science, 28(3), 375 –387.
Cooil, B., Keiningham, T. L., Aksoy, L., & Hsu, M. (2007). A longitudinal analysis of
customer satisfaction and share of wallet: investigating the moderating effect
of customer characteristics. Journal of Marketing, 71(1), 67–83.
Cyr, D. (2008). Modeling web site design across cultures: relationships to trust,
satisfaction, and e-loyalty. Journal of Management Information Systems,
24(4), 47–72.
Davis, F. (1989). Perceived usefulness, perceived ease of use, and user acceptance
of information technology. MIS Quarterly, 13(3), 319–340.
Davis-Sramek, B., Droge, C., Mentzer, J., & Myers, M. (2009). Creating commitment
and loyalty behavior among retailers: what are the roles of service quality
Authors / Viability of Cloud Services
43
and satisfaction? Journal of the Academy of Marketing Science, 37(4), 440–
454.
Dellarocas, C. (2003). The Digitization of word of mouth: Promise and challenges of
online feedback mechanisms. Management Science, 49(10), 1407–1424.
Deloitte (2009). Cloud computing - Forecasting change. Retrieved January 31,
2013, from https://www.deloitte.com/assets/Dcom-Global/Local%20Assets
/Documents/TMT/cloud_-_market_overview_and_perspective.pdf
Devaraj, S., Fan, M., & Kohli, R. (2002). Antecedents of B2C channel satisfaction
and preference: validation e-commerce metrics. Information Systems
Research, 13(3), 316–333.
Dichter, E. (1966). How word-of-mouth advertising works. Harvard Business
Review, 44(6), 147–166.
Dodds, W. B., Monroe, K. B., & Grewal, D. (1991). Effects of price, brand, and store
information on buyers’ product evaluations. Journal of Marketing Research,
28(3), 307–319.
Dong, S., Ding, M., Grewal, R., & Zhao, P. (2011). Functional forms of the
satisfaction–loyalty relationship. International Journal of Research in
Marketing, 28(1), 38–50.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with
unobservable variables and measurement error. Journal of Marketing
Research, 18(1), 39–50.
Fullerton, G. (2003). When does commitment lead to loyalty? Journal of Service
Research, 5(4), 333–344.
Authors / Viability of Cloud Services
44
Ganesh, J., Arnold, M. J., & Reynolds, K. E. (2000). Understanding the customer
base of service providers: an examination of the differences between
switchers and stayers. Journal of Marketing, 64(3), 65–87.
Garnefeld, I., Helm, S., & Eggert, A. (2010). Walk your talk: an experimental
investigation of the relationship between word of mouth and communicators’
loyalty. Journal of Service Research, 93–107.
Gefen, D., Rigdon, E., & Straub, D. (2011). An update and extension to SEM
guidelines for administrative and social science research. MIS Quarterly,
35(2), iii–A7.
Gelbrich, K., & Roschk, H. (2011). A meta-analysis of organizational complaint
handling and customer responses. Journal of Service Research, 14(1), 24–
43.
Gittell, J. H. (2002). Relationships between service providers and their impact on
customers. Journal of Service Research, 4(4), 299–311.
Goodhue, D. L., Lewis, W., & Thompson, R. (2006). PLS, small sample size, and
statistical power in MIS Research. In HICSS 2006 Proceedings. Hawaii,
USA.
Goodhue, D. L., Lewis, W., & Thompson, R. (2012). Does PLS have advantages for
small sample size or non-normal data? MIS Quarterly, 36(3), 981–1001.
Gremler, D. D., & Gwinner, K. P. (2000). Customer-employee rapport in service
relationships. Journal of Service Research, 3(1), 82 –104.
Gustafsson, A., & Johnson, M. D. (2004). Determining attribute importance in a
service satisfaction model. Journal of Service Research, 7(2), 124–141.
Authors / Viability of Cloud Services
45
Hair, J. F., Hult, G. T., Ringle, C. M., & Sarstedt, C. M. (2014). A Primer On Partial
Least Squares Structural Equation Modeling (PLS-SEM). Thousand Oaks,
CA, USA: SAGE Publications.
Han, X., Kwortnik, R. J., & Wang, C. (2008). Service loyalty: an integrative model
and examination across service contexts. Journal of Service Research,
11(1), 22–42.
Harris, J., Ives, B., & Junglas, I. (2012). IT consumerization: when gadgets turn into
enterprise IT tools. MIS Quarterly Executive, 11(3), 99–112.
Heitmann, M., Lehmann, D. R., & Herrmann, A. (2007). Choice goal attainment and
decision and consumption satisfaction. Journal of Marketing Research,
44(2), 234–250.
Hennig-Thurau, T., Gwinner, K. P., & Gremler, D. D. (2002). Understanding
relationship marketing outcomes: An integration of relational benefits and
relationship quality. Journal of Service Research, 4(3), 230–247.
Henseler, J., Ringle, C. M., & Sinkovics, R. (2009). The use of partial least squares
path modeling in international marketing. In S. Zou (Ed.), Advances in
International Marketing (Vol. 20, pp. 277–319). Bingley, England: Emerald
Group Publishing Limited.
Hess, T. J., Fuller, M. A., & Mathew, J. (2006). Involvement and decision-making
performance with a decision aid: The influence of social multimedia, gender,
and playfulness. Journal of Management Information Systems, 22(3), 15–54.
Hinkin, T. R., & Tracey, J. B. (1999). An analysis of variance approach to content
validation. Organizational Research Methods, 2(2), 175–186.
Authors / Viability of Cloud Services
46
Homburg, C., & Fürst, A. (2005). How organizational complaint handling drives
customer loyalty: an analysis of the mechanistic and the organic approach.
Journal of Marketing, 69(3), 95–114.
Homburg, C., Fürst, A., & Prigge, J.-K. (2010). A customer perspective on product
eliminations: how the removal of products affects customers and business
relationships. Journal of the Academy of Marketing Science, 38(5), 531–549.
Homburg, C., Koschate, N., & Hoyer, W. D. (2005). Do satisfied customers really
pay more? A Study of the relationship between customer satisfaction and
willingness to pay. Journal of Marketing, 69(2), 84–96.
Homburg, C., Wieseke, J., & Hoyer, W. D. (2009). Social identity and the service–
profit chain. Journal of Marketing, 73(2), 38–54.
Hong, S., Thong, J. Y. L., & Tam, K. Y. (2006). Understanding continued Information
technology usage behavior: A comparison of three models in the context of
mobile internet. Decision Support Systems, 42(3), 1819–1834.
Iyer, B., & Henderson, J. (2010). Preparing for the future: Understanding the seven
capabilities of cloud computing. MIS Quarterly Executive, 9(2), 117–131.
Jin, Y., & Su, M. (2009). Recommendation and repurchase intention thresholds: a
joint heterogeneity response estimation. International Journal of Research in
Marketing, 26(3), 245–255.
Johnson, M. D., Anderson, E. W., & Fornell, C. (1995). Rational and adaptive
performance expectations in a customer satisfaction framework. Journal of
Consumer Research, 21(4), 695–707.
Johnson, M. W., Christensen, C. M., & Kagermann, H. (2008). Reinventing your
business model. Harvard Business Review, 86(12), 50–59.
Authors / Viability of Cloud Services
47
Katz, M. L., & Shapiro, C. (1986). Technology adoption in the presence of network
externalities. Journal of Political Economy, 94(4), 822–841.
Keiningham, T. L., Cooil, B., Andreassen, T. W., & Aksoy, L. (2007). A longitudinal
examination of net promoter and firm revenue growth. Journal of Marketing,
71(3), 39–51.
Kim, D. J., Ferrin, D. L., & Rao, H. R. (2009). Trust and satisfaction, two stepping
stones for successful e-commerce relationships: a longitudinal exploration.
Information Systems Research, 20(2), 237–257.
Kim, J., Lee, J., Han, K., & Lee, M. (2002). Businesses as buildings: metrics for the
architectural quality of internet businesses. Information Systems Research,
13(3), 239–254.
Kim, S. S., & Son, J.-Y. (2009). Out of dedication or constraint? A dual model of
post-adoption phenomena and its empirical test in the context of online
services. MIS Quarterly, 33(1), 49–70.
Krishnamurthi, L., & Raj, S. P. (1991). An empirical analysis of the relationship
between brand loyalty and consumer price elasticity. Marketing Science,
10(2), 172–183.
Lam, S. Y., Shankar, V., Erramilli, M. K., & Murthy, B. (2004). Customer value,
satisfaction, loyalty, and switching costs: An illustration from a business-to-
business service context. Journal of the Academy of Marketing Science,
32(3), 293–311.
Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in
cross-sectional research designs. Journal of Applied Psychology, 86(1),
114–121.
Authors / Viability of Cloud Services
48
Luo, X., & Homburg, C. (2007). Neglected outcomes of customer satisfaction.
Journal of Marketing, 71(2), 133–149.
MacKenzie, S. B., Podsakoff, P. M., & Paine, J. B. (1999). Do citizenship behaviors
matter more for managers than for salespeople? Journal of the Academy of
Marketing Science, 27(4), 396–410.
MacKenzie, S. B., Podsakoff, P. M., & Podsakoff, N. P. (2011). Construct
measurement and validation procedures in MIS and behavioral research:
Integrating new and existing techniques. MIS Quarterly, 35(2), 293–334.
Malhotra, N. K., Kim, S. S., & Patil, A. (2006). Common method variance in IS
research. Management Science, 52(12), 1865–1883.
Mangold, W. G., Miller, F., & Brockway, G. R. (1999). Word-of-mouth
communication in the service marketplace. Journal of Services Marketing,
13(1), 73–89.
Marcoulides, G., Chin, W. W., & Saunders, C. (2012). When imprecise statistical
statements become problematic: A response to Goodhue, Lewis, and
Thompson. MIS Quarterly, 36(3), 717–728.
Maxham III, J. G., & Netemeyer, R. G. (2002). A longitudinal study of complaining
customers’ evaluations of multiple service failures and recovery efforts.
Journal of Marketing, 66(4), 57–71.
Maxham III, J. G., & Netemeyer, R. G. (2003). Firms reap what they sow: the effects
of shared values and perceived organizational justice on customers’
evaluations of complaint handling. Journal of Marketing, 67(1), 46–62.
Authors / Viability of Cloud Services
49
Mell, P., & Grance, T. (2011). The NIST Definition of Cloud Computing. Retrieved
June 26, 2012, from http://csrc.nist.gov/publications/nistpubs/800-
145/SP800-145.pdf
Miller, K. M., Hofstetter, R., Krohmer, H., & Zhang, Z. J. (2011). How should
consumers’ willingness to pay be measured? An empirical comparison of
state-of-the-art approaches. Journal of Marketing Research, 48(1), 172–184.
Mooi, E., & Sarstedt, M. (2011). A concise guide to market research. New York,
USA: Springer.
Moore, G. C., & Benbasat, I. (1991). Development of an instrument to measure the
perceptions of adopting an information technology innovation. Information
Systems Research, 2(3), 192–222.
Morgeson, F., Mithas, S., Keiningham, T., & Aksoy, L. (2011). An investigation of the
cross-national determinants of customer satisfaction. Journal of the
Academy of Marketing Science, 39(2), 198–215.
Mukherjee, A., & Hoyer, W. D. (2001). The effect of novel attributes on product
evaluation. Journal of Consumer Research, 28(3), 462–472.
Needleman, S. E., & Loten, A. (2012). When freemium fails. The Wall Street
Journal. Retrieved January 12, 2012, from http://online.wsj.com/
news/articles/SB10000872396390443713704577603782317318996
Niculescu, M. F., & Wu, D. J. (2014). Economics of free under perpetual licensing:
Implications for the software industry. Information Systems Research, 25(1),
173–199.
Nijssen, E., Singh, J., Sirdeshmukh, D., & Holzmüeller, H. (2003). Investigating
industry context effects in consumer-firm relationships: preliminary results
Authors / Viability of Cloud Services
50
from a dispositional approach. Journal of the Academy of Marketing Science,
31(1), 46 –60.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric Theory (3rd ed.). Ohio, USA:
McGraw-Hill.
Nyer, P. U. (1997). A study of the relationships between cognitive appraisals and
consumption emotions. Journal of the Academy of Marketing Science, 25(4),
296–304.
Oliva, T. A., Oliver, R. L., & MacMillan, I. C. (1992). A catastrophe model for
developing service satisfaction strategies. Journal of Marketing, 56(3), 83–
95.
Oliver, R. (1977). Effect of expectation and disconfirmation on postexposure product
evaluations: an alternative interpretation. Journal of Applied Psychology,
62(4), 480–486.
Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of
satisfaction decisions. Journal of Marketing Research, 17(4), 460–469.
Oliver, R. L. (1993). Cognitive, affective, and attribute bases of the satisfaction
response. Journal of Consumer Research, 20(3), 418–430.
Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63(4), 33–44.
Olsen, L. L., & Johnson, M. D. (2003). Service equity, satisfaction, and loyalty: from
transaction-specific to cumulative evaluations. Journal of Service Research,
5(3), 184 –195.
Olsen, S. O. (2002). Comparative evaluation and the relationship between quality,
satisfaction, and repurchase loyalty. Journal of the Academy of Marketing
Science, 30(3), 240–249.
Authors / Viability of Cloud Services
51
Otim, S., & Grover, V. (2006). An empirical study on web-based services and
customer loyalty. European Journal of Information Systems, 15(6), 527–541.
Palmatier, R. W., Scheer, L. K., & Steenkamp, J.-B. E. . (2007). Customer loyalty to
whom? managing the benefits and risks of salesperson-owned loyalty.
Journal of Marketing Research, 44(2), 185–199.
Pavlou, P. A., Liang, H., & Xue, Y. (2007). Understanding and mitigating uncertainty
in online exchange relationships: a principal-agent perspective. MIS
Quarterly, 31(1), 105–136.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common
method biases in behavioral research: a critical review of the literature and
recommended remedies. Journal of Applied Psychology, 88(5), 879–903.
Raimondo, M. A., “Nino” Miceli, G., & Costabile, M. (2008). How relationship age
moderates loyalty formation. Journal of Service Research, 11(2), 142 –160.
Rao, L. (2010). Ning Goes Premium. TechCrunch. Retrieved May 15, 2014, from
http://techcrunch.com/2010/05/04/ning-goes-premium/
Ray, S., Kim, S. S., & Morris, J. G. (2012). Research note - online users’ switching
costs: their nature and formation. Information Systems Research, 23(1),
197–213.
Redshift Research. (2012). 45% are unsure where data is hosted. Retrieved
January 24, 2013, from https://www.q3internet.co.uk/press-releases/45-are-
unsure-where-data-is-hosted/
Reichheld, F. F. (2003). The one number you need to grow. Harvard Business
Review, 81(12), 46–54.
Authors / Viability of Cloud Services
52
Reichheld, F. F., & Sasser, W. E. J. (1990). Zero defections: quality comes to
services. Harvard Business Review, (68), 105–111.
Reimann, D. (2013). Switching to Freemium Lessons learned from changing
iOctocat’s price model. Retrieved June 12, 2014, from
http://dennisreimann.de/blog/switch-to-freemium/
Ringle, C. M., Sarstedt, M., & Straub, D. (2012). A critical look at the use of PLS-
SEM in MIS Quarterly. MIS Quarterly, 36(1), iii–xiv.
Rossbach, C., & Welz, B. (2011). Survival of the fittest - How Europe can assume a
leading role in the cloud. Retrieved June 20, 2012, from
http://www.rolandberger.com/media/pdf/Roland_Berger_Cloud_Ecosystem_
D_20111130.pdf
Szymanski, D., & Henard, D. (2001). Customer satisfaction: A meta-analysis of the
empirical evidence. Journal of the Academy of Marketing Science, 29(1), 16–
35.
Teece, D. J. (2010). Business models, business strategy and innovation. Long
Range Planning, 43(2–3), 172–194.
Trenz, M., Huntgeburth, J., & Veit, D. (2013). The role of uncertainty in cloud
computing continuance: Antecedents, mitigators, and consequences. In 21st
European Conference on Information Systems (ECIS). Utrecht, Netherlands.
Van Doorn, J., & Verhoef, P. C. (2008). Critical incidents and the impact of
satisfaction on customer share. Journal of Marketing, 72(4), 123–142.
Venters, W., & Whitley, E. A. (2012). A critical review of cloud computing:
researching desires and realities. Journal of Information Technology, 27(3),
179–197.
Authors / Viability of Cloud Services
53
Vock, M., van Dolen, W., & de Ruyter, K. (2013). Understanding willingness to pay
for social network sites. Journal of Service Research, 16(3), 311–325.
Von Wangenheim, F., & Bayón, T. (2007). The chain from customer satisfaction via
word-of-mouth referrals to new customer acquisition. Journal of the Academy
of Marketing Science, 35(2), 233–249.
Walsh, G., Hennig-Thurau, T., Sassenberg, K., & Bornemann, D. (2010). Does
relationship quality matter in e-services? A comparison of online and offline
retailing. Journal of Retailing & Consumer Services, 17(2), 130–142.
Webster, J., & Watson, R. T. (2002). Analyzing the past to prepare for the future:
Writing a literature review. MIS Quarterly, 26(2), xiii – xxiii.
Westbrook, R. A. (1987). Product/ consumption-based affective responses and
postpurchase processes. Journal of Marketing Research, 24(3), 258–270.
Yim, C. K. (Bennett), Tse, D. K., & Chan, K. W. (2008). Strengthening customer
loyalty through intimacy and passion: roles of customer–firm affection and
customer–staff relationships in services. Journal of Marketing Research,
45(6), 741–756.
Zetta. (2010). Cloud Storage Adoption Trends, Barriers and Expectations - a
Survey. Retrieved May 15, 2012, from http://pages.zetta.net/rs/zetta/
images/Zetta_Cloud_Storage_Survey.pdf
Zhang, J., & Bloemer, J. M. M. (2008). The impact of value congruence on
consumer-service brand relationships. Journal of Service Research, 11(2),
161–178.
Authors / Viability of Cloud Services
54
Zhang, Q., Cheng, L., & Boutaba, R. (2010). Cloud computing: state-of-the-art and
research challenges. Journal of Internet Services and Applications, 1(1), 7–
18.
Authors / Viability of Cloud Services
55
Appendix A: Literature Analysis
Methodology of the literature review
We conducted a structured literature review to create a full picture of previous
literature on the relationships between our four performance indicators satisfaction,
loyalty, word-of-mouth and willingness to pay. Our search was conducted within the
AIS Senior Basket of top IS journals as well as the top journals in marketing and
service research.
Search space (in alphabetical order): European Journal of Information Systems
(EJIS), Information Systems Journal (ISJ), Information Systems Research (ISR),
International Journal of Research in Marketing (IJRS), Journal of AIS (JAIS), Journal
of Consumer Research (JCR), Journal of Information Technology (JIT), Journal of
Marketing (JM), Journal of Marketing Research (JMR), Journal of MIS (JMIS),
Journal of Service Research (JSR), Journal of Strategic Information Systems (JSIS),
Journal of the Academy of Marketing Science (JAMS), Marketing Science (MS), MIS
Quarterly (MISQ)
Selection criteria: At least two of our four major constructs are included in a
quantitative study.
Overall, 62 papers have been found. We used a concept matrix (Webster & Watson,
2002) to structure our findings. Besides the results of the tested relationships in
each paper, we line out type of satisfaction, reference point of loyalty, context and
examination object compared to our study in the table below.
56
Reference Journal
Relationships tested
WTP Type Loyalty Type Satisfaction
Type Context and examination object
SAT →
LOY
SAT →
WOM
LOY →
WOM
SAT →
WTP
LOY →
WTP
WTP →
WOM
Our Study JAIS +* +* +* -*/ns +*/+* Upgrade/ Retention
Vendor Cumulative B2C: Cloud storage services
Agustin & Singh (2005)
JMR ns n/a Vendor Cumulative B2C: Consumer Goods + Offline Service
Andreassen (1999) +* n/a Vendor Transactional B2C: Offline Service
Arens & Rust (2012) JAMS +* n/a Vendor Cumulative B2C: Service (not specified)
Beatty et al. (2012) JSR +* n/a Vendor n/a B2C: Service (not specified)
Blocker et al. (2011) JAMS +* n/a Product Cumulative B2B: ICT Service
Boenigk and Helmig (2013)
JSR +* n/a Vendor Cumulative B2C: Non-profit
Brady et al. (2012) JM +* +* n/a Vendor Transaction-specific
B2C: Offline Service
Brakus et al. (2009) JM +* n/a Vendor Cumulative B2C: Consumer Goods
Brown et al. (2005) JAMS +* +* +* n/a n/a Cumulative B2C: Consumer Goods + Offline Service
Note: * = significant relationship found; ns = relationship not significant; B2C = business-to-consumer; B2B = business-to-business; EJIS = European Journal of Information Systems; IJRM = International Journal of Research in Marketing; ISR = Information Systems Research; JAMS = Journal of the Academy of Marketing Science; JM = Journal of Marketing; JMR = Journal of Marketing Research; JSR = Journal of Service Research; MISQ = Management Information Systems Quarterly.
Authors / Viability of Cloud Services
57
Reference Journal
Relationships tested
WTP Type Loyalty Type Satisfaction
Type Context and examination object
SAT →
LOY
SAT →
WOM
LOY →
WOM
SAT →
WTP
LOY →
WTP
WTP →
WOM
Our Study JAIS +* +* +* -*/ns +*/+* Upgrade/ Retention
Vendor Cumulative B2C: Cloud storage services
Chandrashekaran et al. (2007)
JMR +* n/a Vendor Cumulative B2B: Service (not specified)
Chiou & Droge (2006)
JAMS +* n/a Vendor Cumulative B2C: Consumer Goods
Chiou et al. (2002) JSR +* +* n/a Vendor Cumulative B2C: Offline Service
Chitturi et al. (2008) JM +* +* n/a Product Transaction-specific
B2C: Consumer Goods
Colgate & Danaher (2000)
JAMS +* +* n/a Vendor Cumulative B2C: Offline Service
Cooil et al. (2007) JM +* n/a Vendor Cumulative B2C: Offline Service
Cyr (2008) JMIS +* n/a Vendor Transaction-specific
B2C: Website
Davis-Sramek et al. (2009)
JAMS +* n/a Vendor Cumulative B2B: Industrial Goods
Dong et al. (2011) IJRM +* n/a Product Transaction-specific
B2B/B2C: Consumer Goods
Eisingerich et al. (2013)
JSR +* +* +* ns Classical Vendor Cumulative B2C: Banking
Evanschitzky et al. (2012)
JAMS +* +* Retention Vendor/ Program
Cumulative B2C: Retailing
Note: * = significant relationship found; ns = relationship not significant; B2C = business-to-consumer; B2B = business-to-business; EJIS = European Journal of Information Systems; IJRM = International Journal of Research in Marketing; ISR = Information Systems Research; JAMS = Journal of the Academy of Marketing Science; JM = Journal of Marketing; JMR = Journal of Marketing Research; JSR = Journal of Service Research; MISQ = Management Information Systems Quarterly.
Authors / Viability of Cloud Services
58
Reference Journal
Relationships tested
WTP Type Loyalty Type Satisfaction
Type Context and examination
object
SAT →
LOY
SAT →
WOM
LOY →
WOM
SAT →
WTP
LOY →
WTP
WTP →
WOM
Our Study JAIS +* +* +* -*/ns +*/+* Upgrade/ Retention
Vendor Cumulative B2C: Cloud storage services
Fullerton (2003) JSR +* Retention Vendor n/a B2C: ICT Service
Ganesh et al. (2000) JM +* n/a Vendor Cumulative B2C: Offline Service
Garnefeld et al. (2010)
JSR n/a Vendor n/a B2C: ICT Service
Gittell (2002) JSR +* n/a n/a Cumulative B2C: Offline Service
Gong et al. (2013) JSR +* n/a Vendor Cumulative B2B: Employees of call centers
Gremler & Gwinner (2000)
JSR +* +* +* n/a Vendor Cumulative B2C: Offline Service
Gustafsson & Johnson (2004)
JSR * n/a Vendor Cumulative B2C: Offline Service
Han et al. (2008) JSR * n/a Vendor Cumulative B2C: Offline Service
Heitmann et al. (2007)
JMR +* +* +* n/a Product Transaction-specific
B2C: Consumer Goods
Hennig-Thurau et al. (2002)
JSR +* +* n/a Vendor Cumulative B2C: Offline Service
Homburg & Fürst (2005)
JM ns n/a Vendor Cumulative B2C/B2B: Not specified
Note: * = significant relationship found; ns = relationship not significant; B2C = business-to-consumer; B2B = business-to-business; EJIS = European Journal of Information Systems; IJRM = International Journal of Research in Marketing; ISR = Information Systems Research; JAMS = Journal of the Academy of Marketing Science; JM = Journal of Marketing; JMR = Journal of Marketing Research; JSR = Journal of Service Research; MISQ = Management Information Systems Quarterly.
Authors / Viability of Cloud Services
59
Reference Journal
Relationships tested
WTP Type Loyalty Type Satisfaction
Type Context and examination
object
SAT →
LOY
SAT →
WOM
LOY →
WOM
SAT →
WTP
LOY →
WTP
WTP →
WOM
Our Study JAIS +* +* +* -*/ns +*/+* Upgrade/ Retention
Vendor Cumulative B2C: Cloud storage services
Homburg et al. (2005)
JM -* Classical Product Cumulative + TA-specific
B2C: Offline Service + Consumer Goods
Homburg et al. (2009)
JM +* ns Classical Vendor Transaction-specific
B2C: Offline Service
Homburg et al. (2010)
JAMS +* n/a Product Transaction-specific
B2B: not specified
Jha et al. (2013) JSR +* n/a Vendor Cumulative B2C: Banking Services
Jin & Su (2009) IJRM +* n/a Vendor Cumulative B2C: Consumer Goods
Keiningham et al. (2007)
JM n/a Product Transaction-specific
B2C: Service (not specified)
Kim et al. (2002) ISR *+ n/a Vendor Transaction-specific
B2C: E-Commerce
Kim et al. (2009) ISR +* n/a Vendor Cumulative B2C: E-Commerce
Kim & Son (2009) MISQ +* +* Ns n/a Vendor Cumulative B2C: Online Service
Lam et al. (2004) JAMS +* +* n/a n/a Cumulative B2B: Offline Service
Maxham III & Netemeyer (2003)
JM n/a n/a Transaction-specific
B2C: E-Commerce
Note: * = significant relationship found; ns = relationship not significant; B2C = business-to-consumer; B2B = business-to-business; EJIS = European Journal of Information Systems; IJRM = International Journal of Research in Marketing; ISR = Information Systems Research; JAMS = Journal of the Academy of Marketing Science; JM = Journal of Marketing; JMR = Journal of Marketing Research; JSR = Journal of Service Research; MISQ = Management Information Systems Quarterly.
Authors / Viability of Cloud Services
60
Reference Journal
Relationships tested
WTP Type Loyalty Type Satisfaction
Type Context and examination
object
SAT →
LOY
SAT →
WOM
LOY →
WOM
SAT →
WTP
LOY →
WTP
WTP →
WOM
Our Study JAIS +* +* +* -*/ns +*/+* Upgrade/ Retention
Vendor Cumulative B2C: Cloud storage services
Maxham III & Netemeyer (2002)
JM n/a Vendor Cumulative B2C: Offline Service
Morgeson et al. (2011)
JAMS n/a n/a n/a B2C: Not specified
Nijssen et al. (2003) JAMS +* n/a Vendor Cumulative B2C: Consumer Goods + Offline Service
Nyer (1997) JAMS +* n/a n/a Transaction-specific
B2C: Consumer Goods
Oliva et al. (1992) JM +* n/a Vendor Cumulative B2C+B2B: Offline Service
Olsen & Johnson (2003)
JSR +* n/a Vendor Cumulative + TA-specific
B2C: Offline Service
Olsen (2002) JAMS +* n/a Product Cumulative B2C: Consumer Goods
Otim & Grover (2006)
EJIS +* n/a Vendor Transaction-specific
B2C: E-Commerce
Palmatier et al. (2007)
JMR +* Retention Vendor n/a B2B: Industrial Goods
Note: * = significant relationship found; ns = relationship not significant; B2C = business-to-consumer; B2B = business-to-business; EJIS = European Journal of Information Systems; IJRM = International Journal of Research in Marketing; ISR = Information Systems Research; JAMS = Journal of the Academy of Marketing Science; JM = Journal of Marketing; JMR = Journal of Marketing Research; JSR = Journal of Service Research; MISQ = Management Information Systems Quarterly.
Authors / Viability of Cloud Services
61
Reference Journal
Relationships tested
WTP Type Loyalty Type Satisfaction
Type Context and examination
object
SAT →
LOY
SAT →
WOM
LOY →
WOM
SAT →
WTP
LOY →
WTP
WTP →
WOM
Our Study JAIS +* +* +* -*/ns +*/+* Upgrade/ Retention
Vendor Cumulative B2C: Cloud storage services
Raimondo et al. (2008)
JSR +* n/a Vendor Cumulative B2C: ICT Service
Ray et al. (2012) ISR +* n/a Vendor Cumulative B2C: Internet Provider Service
van Doorn & Verhoef (2008)
JM n/a Product Cumulative B2B: Offline Service
von Wangenheim & Bayón (2007)
JAMS +* n/a n/a Cumulative B2C/B2B: Offline Service
Walsh et al. (2010) JAMS Ns n/a Vendor Cumulative B2C: Offline + Online Service
Westbrrok (1987) JMR ns n/a n/a Cumulative B2B: Consumer Goods + Offline Service
Wieseke et al. (2012) JSR +* n/a Vendor Cumulative B2C: Travel agency service
Yim et al. (2008) JMR n/a Product Cumulative B2C: Offline Service
Zhang & Bloemer (2008)
JSR ns +* +* +* +* Retention Vendor Cumulative B2C: Consumer Goods + Offline Service
Note: * = significant relationship found; ns = relationship not significant; B2C = business-to-consumer; B2B = business-to-business; EJIS = European Journal of Information Systems; IJRM = International Journal of Research in Marketing; ISR = Information Systems Research; JAMS = Journal of the Academy of Marketing Science; JM = Journal of Marketing; JMR = Journal of Marketing Research; JSR = Journal of Service Research; MISQ = Management Information Systems Quarterly.
62
Appendix B: Measurement Items for Principal Constructs
Perceived Uncertainty (Pavlou, Liang, & Xue, 2007)
UNC1: I feel that using [cloud service] involves a high degree of uncertainty. UNC2: I feel the uncertainty associated with using [cloud service] is high. UNC3: I am exposed to many uncertainties if I am using [cloud service]. UNC4: There is a high degree of uncertainty when using [cloud service].
Perceived Ease of Use (Davis, 1989; Pavlou et al., 2007)
PEU1: I find [cloud service] easy to use. PEU2: Using [cloud service] does not require a lot of mental effort. PEU3: I find it easy to get [cloud service] to do what I want it to do. PEU4: The use of [cloud service] is clear and understandable.
Perceived Usefulness (Davis, 1989; Pavlou et al., 2007)
Using [cloud service] enhances my effectiveness. Using [cloud service] enhances my productivity. Using [cloud service] improves my performance. Using [cloud service] enables me to accomplish tasks more quickly.
Perceived Value of Upgrade (Mukherjee & Hoyer, 2001)
An upgrade of %cloud service% (more storage and security) is likely...... PVU1: ... to offer me a lot of advantages. PVU2: ... add a lot of value. PVU3: ... increase the benefit of using the service.
Satisfaction (Lam et al., 2004)
SAT1: I am very contented with %cloud service%. SAT2: I am very pleased with %cloud service%. SAT3: Overall, I am very satisfied with %cloud service%. SAT4: Overall, the %cloud service% comes up to my expectations
Word-of-Mouth (S. S. Kim & Son, 2009)
WOM1: Ich werde meine Freunde einladen, [cloud service] zu nutzen. WOM2: Ich werde anderen [cloud service] empfehlen. WOM3: Ich werde Freunde und Bekannte zu [cloud service] einladen. WOM4: Ich werde meinen Kollegen [cloud service] empfehlen.
Loyalty (Ray, Kim, & Morris, 2012)
LOY1: It means a lot to me to continue to use [cloud service]. LOY2: I feel loyal towards [cloud service]. LOY3: I consider myself to be highly loyal to [cloud service]. LOY4: It means a lot to me to remain a customer of [cloud service].
Willingness to Pay for Retention (S. S. Kim & Son, 2009)
Imagine [cloud service] would no longer be freely available. How likely are the following statements? WTPR1: I am willing to pay € 0.50 per month for [cloud service]. WTPR2: I am willing to pay a one-time only fee of € 5 for [cloud service]. WTPR3: I am willing to pay an annual fee of €3 for [cloud service]. WTPR4: I am willing to pay a semi-annually fee of € 1.50 for this service.
Willingness to Pay for Upgrade (Vock et al., 2013)
WTPU1: I am willing to pay a premium for additional services of [cloud service]. WTPU2: I am willing to pay a premium for advanced features (e.g., more storage, better access) of [cloud service]. WTPU3: I am willing to pay a premium for advanced security of [cloud service]. WTPU4: I will upgrade to paid [cloud service] account soon.
Cloud Knowledge (Hess, Fuller, & Mathew, 2006)
CK1: I understand how cloud technology is functioning. CK2: I have a good grasp of how cloud technology works. CK3: I can easily describe the functionality provided by cloud technology. CK4: It is easy for me to recall how cloud technology functions.
IT Experience
ITE1: I know a lot about technology. ITE2: I know a lot about computers. ITE3: I know a lot about the internet. ITE4: I know a lot about cloud services.
Authors / Viability of Cloud Services
63
Appendix C: Correlation Matrix and AVE
LOY PEU PU PVU SAT UNC WOM WTPR WTPU
LOY 0,89 - - - - - - - -
PEU 0,46 0,84 - - - - - - -
PU 0,56 0,58 0,92 - - - - - -
PVU 0,20 0,13 0,26 0,96 - - - - -
SAT 0,60 0,68 0,59 0,18 0,90 - - - -
UNC -0,29 -0,30 -0,22 0,06 -0,37 0,92 - - -
WOM 0,58 0,41 0,51 0,31 0,56 -0,23 0,89 - -
WTPR 0,25 0,16 0,26 0,44 0,21 -0,09 0,31 0,87 -
WTPU 0,16 0,04 0,19 0,42 0,07 0,01 0,21 0,44 0,89
Note: The diagonal elements (in bold) represent the square root of AVE.
Appendix D: Cross and Outerloadings
LOY PEU PU SAT UNC PVU WOM WTPR WTPU
LOY1 0.858 0.410 0.513 0.543 -0.238 0.177 0.567 0.240 0.156
LOY2 0.855 0.416 0.469 0.544 -0.330 0.132 0.431 0.206 0.128
LOY3 0.902 0.380 0.463 0.481 -0.235 0.175 0.471 0.210 0.136
LOY4 0.929 0.415 0.527 0.558 -0.259 0.195 0.570 0.261 0.152
WOM1 0.509 0.346 0.413 0.491 -0.180 0.312 0.918 0.318 0.214
WOM2 0.552 0.423 0.471 0.570 -0.267 0.250 0.895 0.270 0.178
WOM3 0.497 0.332 0.408 0.446 -0.154 0.276 0.893 0.278 0.197
WOM4 0.498 0.342 0.485 0.461 -0.191 0.298 0.853 0.262 0.193
WTPU1 0.164 0.039 0.187 0.086 -0.041 0.378 0.196 0.389 0.922
WTPU2 0.164 0.027 0.181 0.083 -0.026 0.382 0.204 0.406 0.920
WTPU3 0.117 0.051 0.172 0.050 0.093 0.412 0.211 0.411 0.827
WTPU4 0.133 0.010 0.152 0.046 -0.013 0.365 0.165 0.366 0.888
WTPR2 0.239 0.212 0.270 0.244 -0.095 0.363 0.293 0.831 0.364
WTPR3 0.223 0.104 0.234 0.182 -0.082 0.411 0.273 0.894 0.393
WTPR4 0.212 0.074 0.206 0.148 -0.046 0.398 0.253 0.889 0.401
PVU1 0.201 0.120 0.267 0.177 0.044 0.957 0.299 0.423 0.395
PVU2 0.175 0.140 0.285 0.181 0.047 0.954 0.323 0.445 0.433
PVU3 0.178 0.116 0.218 0.143 0.058 0.960 0.291 0.416 0.414
SAT1 0.541 0.625 0.520 0.915 -0.367 0.123 0.502 0.205 0.055
SAT2 0.6 0.608 0.579 0.895 -0.323 0.207 0.539 0.232 0.086
SAT3 0.566 0.622 0.545 0.929 -0.336 0.149 0.531 0.215 0.081
SAT4 0.434 0.577 0.424 0.849 -0.293 0.144 0.409 0.150 0.043
PEU1 0.394 0.923 0.524 0.639 -0.237 0.124 0.365 0.136 0.034
PEU2 0.253 0.720 0.242 0.425 -0.253 0.038 0.234 0.119 -0.013
PEU3 0.458 0.817 0.594 0.565 -0.230 0.146 0.399 0.145 0.055
PEU4 0.424 0.916 0.545 0.638 -0.268 0.121 0.366 0.136 0.037
PU1 0.535 0.556 0.930 0.568 -0.194 0.231 0.483 0.267 0.163
PU2 0.509 0.522 0.933 0.524 -0.198 0.279 0.468 0.260 0.201
PU3 0.495 0.528 0.913 0.539 -0.205 0.266 0.451 0.256 0.183
PU4 0.514 0.515 0.901 0.499 -0.200 0.210 0.435 0.231 0.170
UNC1 -0.264 -0.231 -0.189 -0.297 0.870 0.059 -0.186 -0.078 -0.020
UNC2 -0.247 -0.243 -0.147 -0.306 0.901 0.063 -0.180 -0.064 0.033
UNC3 -0.274 -0.284 -0.216 -0.366 0.950 0.045 -0.218 -0.101 0.009
UNC4 -0.303 -0.291 -0.232 -0.371 0.941 0.028 -0.234 -0.076 -0.002