You are on page 1of 8

Adoption of Cloud Computing by South African Firms: The

Role of Institutional Forces, Absorptive Capacity, and Top


Management
Jason F Cohen Jian Mou Jonathan Trope
Wits University Wits University Wits University
Private Bag X3 Private Bag X3 Private Bag X3
WITS, 2050, South Africa WITS, 2050, South Africa WITS, 2050, South Africa
+27 11 717 8164 +27 11 717 8160 +27 11 717 8160
jason.cohen@wits.ac.za jian.mou@wits.ac.za jtrope@gmail.com

ABSTRACT these cloud services enable individuals and organisations to


Cloud computing offers a number of advantages for organisations rapidly connect to data, provision IT infrastructure, and access
looking to rent on-demand IT resources as a service. However, sophisticated applications [1].
there are also uncertainties and challenges to cloud computing use Cloud computing can make the latest software and hardware
that are tempering the rate of adoption. This paper contributes to readily available to organisations that may previously not have
our understanding of cloud computing adoption within South had the resources to own or manage their own IT assets [2]. By
African firms by developing and testing a research model. Our providing for access to remote IT resources on a pay-by-use basis,
model hypothesizes mimetic and normative institutional cloud computing can also lead to capital cost reductions and
pressures, top management support, and the absorptive capacity eliminate barriers to entry associated with the purchase of
of the firm as factors influencing cloud computing adoption. Data expensive IT resources [3, 4]. Cloud computing also allows for
collected from 87 South African firms confirmed most rapid provisioning of IT services within flexible timeframes [5].
hypothesized relationships, and supported institutional and For example, an organisation that has an urgent requirement for
absorptive capacity theories as explanations for adoption additional back-up storage capacity could be provided this
behavior. Top management support was found to be the most capacity via a cloud service without the need to physically
important factor influencing adoption and also partially mediated acquire and install the storage technology. Cloud computing also
the effects of institutional pressures. offers scalability, which allows for IT resources to be ramped up
or down dynamically based on requirements. For example, if
Categories and Subject Descriptors transaction volumes are exceeding a server’s capacity then
K.6.0 [Management of Computing and Information Systems]: additional computing capacity can be dynamically allocated to the
General server. Cloud computing also makes vast IT resources available
for processing-intensive applications by allowing for additional
processing power to be utilised when required [6]. Further
General Terms benefits of cloud computing include the availability of backup
Management, Theory. systems, disaster recovery, and resiliency to infrastructure failures
which are available from the cloud service provider [2].
Keywords Furthermore, cloud computing offers an alternative for
Cloud computing, technology adoption, institutional theory, organisations wanting to outsource their non-core or
absorptive capacity, top management. commoditised IT operations [4]. For example, an organisation
may choose to move to a SaaS model for office productivity tools
and hosted email. Cloud service offerings are thus said to
1. INTRODUCTION introduce efficiencies that are often difficult to replicate in-house
Cloud computing has emerged as a high-potential IT innovation [7], and are particularly promising for those organisations
that replaces the need for organisationally owned IT assets with struggling to keep up with high rates of technlology change and
on-demand and rapidly scalable alternatives delivered as services who are under financial pressures from continued investment in
over the Internet. There are three delivery models for cloud in-house software and hardware [8].
computing services: Software-as-a-service (SaaS), Platform-as-a-
service (PaaS) and Infrastructure-as-a-service (IaaS). Together, Cloud computing is not however without its challenges. Lin and
Chen [9] identify the lack of formalised standards across cloud
Permission to make digital or hard copies of all or part of this work for providers as a major concern because of the implications for
personal or classroom use is granted without fee provided that copies are
interoperability and portability of data when moving between
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. Copyrights
providers. It is also possible that some systems within an
for components of this work owned by others than ACM must be honored. organisation might not be compatible with a cloud computing
Abstracting with credit is permitted. To copy otherwise, or republish, to model but might nevertheless need to integrate with other cloud-
post on servers or to redistribute to lists, requires prior specific based systems, making these interactions and management thereof
permission and/or a fee. Request permissions from technologically complex [6]. Given that cloud computing allows
Permissions@acm.org. for IT resources to be distributed, the IT skills and knowledge
SAICSIT2104, September 29 - October 01 2014, Centurion, South Africa
Copyright 2014 ACM 978-1-4503-3246-0/14/09…$15.00
http://dx.doi.org/10.1145/2664591.2664604n

30
required to effectively manage and control these distributed IT and instead accounts for the influence of the social context of the
resources may not exist within an organisation [10]. Another often organisation [17]. The motivation to conform is driven by the
cited concern relates to data security where monitoring and need to acquire legitimacy, without which organisations may
enforcement of data security on multi-tenant IT infrastructure is struggle to secure status, resources and other forms of support [18,
far more challenging than in an in-house computing environment 19]. Dimaggioand and Powell [18] propose three pressures
[11]. For an organisation using cloud based systems, e.g. SaaS, a through which conformity occurs: mimetic, coercive and
mobile device such as a laptop or smartphone can act as a normative. Mimetic pressures lead an organisation to change over
gateway to those systems and increase the risk of unintended time in order to become more like other organisations in its
access to sensitive corporate data [5]. Organisations must also environment [20]. The actions of competitors are a particularly
weigh-up cloud adoption in relation to existing hardware and relevant source of mimetic pressure. Coercive pressures to
software investments and the long-term consequences of conform originate from environmental sources such as from a
discontinuing their direct ownership of IT assets. dominant customer or supplier, or from government regulations.
Normative pressures result from the presence of dyadic
With the advantages and benefits that cloud computing
relationships amongst members of a network that leads to
introduces, it is a technology that could hold much promise for
consensus and the development of norms and rules that become
firms yet there are clear obstacles to more widespread adoption [9,
entrenched amongst members of a network.
10, 12]. Prior empirical research has been undertaken in an effort
to better understand adoption of cloud computing (e.g. [6, 8, 13]). Institutional theory is a promising theoretical lens from which to
However such studies have not specifically focused on understand the adoption and diffusion of IT in organisational
developing countries such as South Africa where cloud computing contexts [15, 21]. In the context of IT adoption, mimetic pressures
is still in its infancy in terms of awareness and adoption [11]. have are seen to influence organisations to imitate the adoption
Moreover, prior studies have often not been underpinned by behaviours of leading peers as a response to uncertainty regarding
theory, and have not sufficiently considered how adoption might the promise of an IT innovation. Coercive pressures to adopt
be affected by inter-relationships amongst external institutional technologies include compliance pressures exerted on firms by
influences arising from the organisation’s social context, internal government regulation and/or by other dominant organisations
organisational factors such as absorptive capacity and top (e.g. a dominant customer) on which they are dependent. Lastly,
management support, and technological factors including the normative pressures influence adoption by establishing
complexity of the innovation. Understanding how such factors professional norms and widespread agreement on new
interact and combine to influence adoption of new technologies technologies that are shared between organisations in a relational
remains of continued theoretical and practical significance. network (see for example [20]). Empirical results from past IT
adoption studies suggest that mimetic, coercive and normative
The purpose of our study is to address this gap. We extend the pressures are generally important to organisational adoption [21].
work of [14, 15, 16] into adoption of IT innovations by drawing
However, the relative effects of these pressures are likely to differ
inter-alia on Institutional Theory and Absorptive Capacity Theory
across different technology adoption contexts. For example,
as theoretical lenses from which to develop a research model of
Khalifa and Davison [22] showed coercive pressures to be
cloud computing adoption by South African organisations. We
important in the context of Electronic Trading Systems adoption,
collected data using a questionnaire instrument administered to IT
but Bharati et al. [23] could not show such pressures important in
decision makers within a sample of South African firms. The the context of organisational social media adoption. Similarly,
model is tested using the partial least squares (PLS) approach to
Teo et al. [20] showed mimetic pressures important in the context
structural equation modelling (SEM).
of financial electronic data interchange adoption, but Liu et al.
Results have important implications for organisations wishing to [24] could not show mimetic pressures important in the context of
understand the factors promoting the adoption of cloud Internet-enabled supply chain management systems adoption.
computing. We demonstrate the important role played by the Therefore, when drawing on institutional theory, it’s necessary to
social context of the organization and how external pressures and consider which of the three pressures are likely to be more
a need for legitimacy influences organizational adoption prevalent in a specific adoption context. This study focuses on
behaviours. We also provide important evidence for top managers mimetic and normative institutional pressures. Coercive pressures
on the role they play in responding to external pressures in the are less likely to be important to adoption for technologies not
process of adopting technology innovations. Moreover, we subject to strong government regulation, or technologies not
highlight the importance of the prior knowledge and abilities of specifically enabling inter-organisational operability (e.g. [20,
employees which provide the absorptive capacity organisations 22]). We therefore omit coercive pressures and focus our attention
require to recognise and appreciate the potential opportunity or on the influence on mimetic and normative pressures.
threat that new innovations present.
2.2 Top Management Support
2. THEORETICAL BACKGROUND AND The influence of institutional pressure on an organisation is likely
to be mediated by top management. Top managers - as
RESEARCH MODEL
organisational decision makers - provide the link between the
2.1 Institutional Theory external environment and organisational behaviors i.e.
Institutional Theory suggests that organisational changes, whether institutional pressures are first interpreted by top managers before
structural or behavioural, are driven by pressure on organisations they are acted upon [14]. Top managers with different beliefs in
to conform. The major sources of such pressure are other the relative advantages and organisational changes needed to
organisations. Institutional theory thus moves beyond the simple accommodate institutional pressures are likely to adopt different
economic or rationale perspective on organisational behaviours behaviours [25].

31
In the IT context, the role of top management in the successful 2.4.1 The effects of institutional pressures
adoption and diffusion of IT has long been recognized [26]. Top To reflect both the general pressures to conform to peer
management with strong beliefs regarding benefits will support IT organisations and imitate their actions, as well as the pressure to
efforts by creating a positive environment for implementation observe and imitate the behaviours of organisations that are
including active participation and provision of resources [14]. The perceived to be especially successful, two first-order factors,
adoption of new IT innovations is less likely to occur in those namely adoption of cloud computing by competitors and
organisational contexts where top managers have a muted perceived competitor success with cloud computing are depicted
response to institutional pressures and fail to believe in the in our research model.
adoption effort.
When technologies are poorly understood or when there is
uncertainty regarding returns from an initiative, organisations will
2.3 Absorptive Capacity Theory model their response to these technologies on other organisations
Absorptive capacity is a firm’s ability to identify, assimilate and they perceived to be legitimate and successful [18, 19]. If there
apply external knowledge to its advantage [27]. Absorptive are enough similar organisations that function and behave in a
capacity builds on existing knowledge and abilities of employees certain manner to the extent that functioning in that manner
available within the organisation and is further developed through becomes a legitimate way to function, then others will likely
training, internal research and development programmes, and follow so as to avoid being identified as being less innovative or
repetitive application of internalized knowledge [27, 28]. Without unresponsive [20]. The behaviours of successful organisations
developing absorptive capacity, firm’s constrain their potential to may be considered especially legitimate. Organisations can learn
assimilate knowledge about innovations and may even resist about these behaviours through observation and imitate those
knowledge about new innovations entering the firm. On the other organisational behaviours and practice, or even avoid practices,
hand a firm with stronger absorptive capacity is able to recognise according to their perceived impact or outcomes on the observed
and appreciate the potential opportunity or threat that new organisation [33]. Therefore organisations are more likely to
innovations present [27, 29]. Absorptive capacity thus lowers the adopt cloud computing if they perceive adoption as legitimate and
knowledge barriers surrounding complex technologies and as a contributing factor to peer organisation success. Thus:
enables firms to more successfully deploy them [30]. Prior IS
research has found absorptive capacity important to strategic IT H1: Greater mimetic pressures will lead to greater adoption of
assimilation [31], and organisational adoption of innovations such cloud computing.
as enterprise systems [14], social media [23], and business
Once an IT innovation is available to an industry, members in a
intelligence [32].
relational network collectively evaluate and may promote the
features of the innovation. This evaluation and promotion in-turn
2.4 Research Model and Hypotheses shapes the institutional norms regarding adoption of the
Drawing on the above theoretical perspectives, Figure 1 depicts innovation [14]. Adoption can thus be influenced by the extent
the study’s research model. Our dependent variable is adoption of that an organisation’s trading partners have adopted the
cloud computing, which we conceptualize as the extent to which innovation and communicated their reasons for doing so.
cloud services have been adopted by an organisation as a means Therefore, as an organisation perceives more of its relational
to replace legacy technologies and underpin key operational and partners adopting an innovation, adoption of the IT innovation
management information systems. The model depicts the effects may come to be deemed normatively appropriate for the
of two sources of mimetic pressure, and three sources of organisation [20]. Therefore, organisations that may be
normative pressure. These institutional pressures are theorized to considering adoption of cloud computing may be influenced by
influence adoption both directly and via top management. the extent of adoption among their suppliers and customers with
Absorptive capacity is also depicted as influencing adoption. Our which the organisation has direct ties.
research model controls for the effects of two variables identified
Dyadic relationships are not however the only mechanism
from diffusion of innovation theory, namely compatibility and
through which norms are shared and communicated from
complexity. We also control for organization size (number of
organisation to organization. There are some key institutions that
employees and IT staff). The model’s constructs are described in
exert influence on organisational behaviours with respect to IT
more detail below and associated hypotheses are developed.
innovation adoption, these include standards bodies, and
professional and industry associations [20]. Furthermore,
professional affiliations, conferences and the networking of top
managers is another important avenue through which normative
influences can permeate [14]. The latter in particular may be a
further source of normative pressure in the cloud computing
adoption context. If professional bodies with which organisational
decision-makers participate are active in promoting and
disseminating information regarding cloud computing, the
normative influences are likely to be particularly strong and will
lead such organisations to be more inclined towards adoption.
Taken together, it is hypothesized that:
H2: Greater normative pressures will lead to greater adoption of
cloud computing.
Figure 1. Research model

32
2.4.2 The effects of top management resources to inter-alia purchase innovations and withstand
Top management support starts by creating a positive potential risks such as operational failures resulting from
environment for implementation through their beliefs in the technology failure. Number of organizational employees and IT
potential of the IT innovation to benefit the organization [14]. Top employees are used as proxy measures for organisation size.
management support sends strong signals throughout the
organization and reduces political roadblocks to adoption and 3. RESEARCH METHODOLOGY
implementation efforts [34]. Without top management support it McGregor’s Who Owns Whom Directory (2013) was used to
would be difficult to secure necessary resources and the support construct a sampling frame of South African organisations. The
for organisational changes required by technology adoption [35]. directory includes details of both publically listed and private
Previous meta-analytics studies have confirmed top management companies as well as state holding companies and subsidiaries
support as a highly significant factor influencing organisational IT within South Africa. A total of 980 South African organisations
adoption [36]. It can therefore be hypothesized that: were identified from the directory where details of relevant IT
H3: Greater top management support will lead to greater decision makers and executives were provided. IT decision
adoption of cloud computing makers and executives were considered the most appropriate key
informants regarding cloud computing adoption. They are well
Prior empirical studies also support the mediating influence of top positioned to understand their organisation’s IT resources and
management support on the effects of institutional pressures on technological environment and to report on the organisation’s
various organisational behaviours [14, 23, 37, 38]. Top managers current and future plans regarding IT resources.
are considered to be the organisation’s boundary spanners, and
are thus the focal point for external pressures. It is top managers The identified IT executives within the 980 sampled organisations
who respond to the institutional pressures and translate them into were invited to participate in the study by completing a structured
an organisational response. Institutional pressures must therefore questionnaire. The questionnaire was administered via an online
first influence top management’s beliefs in the relative advantages survey site.
offered by the technology before changes are accommodated
internally [25]. We therefore hypothesize that: 3.1 Measurement
Constructs were operationalized based on previously validated
H4: Top management support for the adoption of cloud
instruments. Cloud computing adoption was measured using three
computing will increase in response to mimetic pressures
items adapted from [14] reflecting the extent to which cloud
H5: Top management support for the adoption of cloud computing was replacing legacy systems and being used to
computing will increase in response normative pressures deliver the technologies that support business operations and
decision making. Items were on a five-point scale from 1=very
2.4.3 The effects of absorptive capacity low to 5=very high. Top management support was measured
Absorptive capacity provides firms with an innovation capability using a 3-item Likert-scale adopted from [14] reflecting top
[27]. Firms with higher levels of absorptive capacity are more management’s belief in the value and importance of using cloud
able to track external changes, reconfigure their resources, and computing. Absorptive capacity is usually measured through
capitalize on new innovations at the opportune time [23, 39]. proxies such as R&D activity or stock of existing knowledge [23,
Differences in absorptive capacity have been used as a means of 27]. We adopted a 4-item Likert scale adopted from [14] to reflect
explaining variations in IT adoption across firms [14]. Thus firms the organisation’s existing stock of knowledge about cloud
with greater existing knowledge and abilities related to cloud computing. One item was however dropped during an initial
computing and who train employees on cloud computing principal components analysis.
technologies will have a greater absorptive capacity to exploit Mimetic pressure was conceptualized as a second-order construct
cloud computing to meet its needs. It is therefore hypothesized formed from two first-order constructs. These were the extent of
that: adoption by competitors (1 item) and the perceived success
H6: Greater absorptive capacity will lead to greater adoption of enjoyed by competitors from adoption (factor score of 4 items).
cloud computing. The items were adopted from [20] and measured on a 5-point
Likert scale.
2.4.4 Controls Normative pressure was also modeled as a formative second-order
We also control for the possible effects on adoption of: construct with three first-order constructs. The first two captured
compatibility and complexity of the innovation, and the extent of cloud computing adoption by suppliers (1 item) and
organisationsize (number of employees and IT staff). the extent of cloud computing adoption by customers (1 item)
each measured on a 5-point scale (1=none to 5=all have adopted).
The complexity of an innovation refers to how difficult it is The third normative pressure for cloud adoption was those arising
perceived to be to understand and use [40]. The greater the from professional and industry bodies (factor score of 3 items).
perceived complexity, the less likely an organization will adopt Items were adopted from [20] and [41].
cloud computing. Moreover, when an innovation is recognised as
being compatible with an organisation’s existing IT, it may be Compatibility and complexity control variables were measured
more likely to be adopted [13]. Cloud computing services might using 3 scale items and 4-scale items respectively [13]. Size was
not be easily adopted in conjunction with existing infrastructure measured as number of organizational employees and number of
and/or applications [9]. IT employees. The instrument was pre-tested and pilot tested with
12 IT managers prior to distribution. Example items are presented
Small organisations are considered less likely to withstand cloud in Table 1.
computing adoption hurdles [14]. Larger organisations have more

33
Table 1. Questionnaire items Organisations of varying sizes were all represented in the sample
with companies having more than 2000 employees being the most
Constructs Operationalization Example items
represented at 38% and smaller organisations with fewer than 100
2nd-order construct Our main competitors
employees, comprising 22% of the total responses. Organisations
with two formative that have adopted
Mimetice with less than 30 IT staff represent just over half of the total at
first-order constructs cloud computing are
Pressures 53%, with organisations having more than 200 IT staff
[20]. perceived favourably
(MIM) representing 24% of the sample.
by others in the same
industry. Table 2. Respondents’ characteristics
2nd-order construct There is a very strong
Demogra- Cate- %
with three formative message in companies
Normative phics gory
first-order constructs that you can’t stay in
Pressures IT Executive 21 24%
[20, 41]. business nowadays if
(NOR) Other (e.g. finance director) 14 16%
you do not adopt cloud
computing. IT Manager 14 16%
3- item scale [14]. The senior CIO 12 14%
management in your Job Titles Technology Executive 7 8%
Top Technology Manager 6 7%
firm believe that cloud
Management General Manager 6 7%
computing has the
Support Operations Executive 4 5%
potential to provide
(TMS) Operations Manager 3 3%
significant business
benefits to the firm. More than 8 37 42%
4- item [14]. 1 item Our company provides 4 to 8 18 21%
Absorptive Years
dropped following regular cloud 2 to 4 18 21%
capacity Employed
PCA. computing training to 1 to 2 9 10%
(ABC)
its IT employees. Less than 1 5 6%
3-item scale adapted Cloud computing is Technology 23 26%
Compatibility from [13]. compatible with my Financial 20 23%
(COM) company’s existing Others 15 17%
hardware and software. Industry Manufacturing 8 9%
4-item scale adapted My company believes type Telecommunications 6 7%
Complexity from [13]. that cloud computing Retail 6 7%
(COP) implementation is a Healthcare 4 5%
complex process. Automotive 4 5%
3-item scale adapted To what extent are More than 30 28 32%
from [14]. cloud computing 20 to 30 12 14%
solutions replacing Company
Adoption 10 to 20 28 32%
legacy technologies Age
5 to 10 12 14%
within your
Less than 5 7 8%
organization?
More than 2000 33 38%
1000 to 2000 6 7%
Number of 500 to 1000 14 16%
4. EMPIRICAL RESULTS Employees 300 to 500 7 8%
100 to 300 8 9%
4.1 Participants Less than 100 19 22%
One hundred and twenty (120) organisations agreed to participate More than 200 21 24%
in the study and completed the survey. However, after removing
100 to 200 3 4%
responses with large amount of missing data, 87 usable responses
Number of 50 to 100 9 10%
remained. Table 2 presents the respondent profiles according to
IT Staff 30 to 50 8 9%
job title, years employed at the organisation, industry type,
10 to 30 18 21%
company age, number of employees, and number of IT staff. The
respondents were all appropriately positioned decision-makers Less than 10 28 32%
with IT responsibilities in their firm.
Respondents with an excess of four year’s employment are very Across the sample it was observed that SaaS is the most diffused
well represented, comprising 63% of the total sample. There was cloud computing service model with 50.6% of the sampled
representation across industry types, with very high representation organisations reporting adoption. IaaS has been adopted by 44.8%
across financial services (23%) and technology industries (26%) of the responding organisations. PaaS is the least adopted cloud
respectively. Companies of various ages are represented in the service model across industries (29.9%). Just under one third of
sample with the vast majority (78%) being well established respondents (28.7%) reported that they have not yet adopted any
organisations at least 10 years or older and 32% being older than of the cloud computing services (Figure 2). We next test our
30 years. research model aimed at explaining the variations in cloud
adoption.

34
(β=0.381, p<0.001) have significant effects on top management
support, which in turn has a significant positive effect on adoption
of could computing (β=0.360, p<0.001). These results confirmed
H1, H3, H4 and H5. However, we cannot confirm the direct
effects of normative pressures on adoption. Thus, H2 is rejected
and the effects of normative pressures are considered fully
mediated by top management. The effect of mimetic pressure on
adoption is partially mediated by top management. Top
management is thus confirmed to playing an important role in
mediating the effects of external pressures on organizational
adoption behaviours. Absorptive capacity also has a significant
effect on adoption (β=0.242, p<0.05). Therefore, H6 is supported.
The control variables were not significant. However, size (number
Figure 2. Cloud Computing Adoption of employees) has a moderate, albeit insignificant, negative effect
on adoption (β=-0.137) suggesting smaller South African firms
4.2 Reliability and Validity are more likely to adopt than larger firms. Perceived complexity
A confirmatory factor analysis using PLS modeling was carried of the technology also had a non-significant negative effect (β=-
out to confirm reliability and validity of the measures. The 0.143) which was in the expected direction. The results are
measurement items all loaded highly onto their expected depicted in Figure 2 and summarized in Table 4.
theoretical constructs. The standardized loadings of the Table 4. Summary of results (* p<.05; ** p<.01; *** p<.001)
measurement items, average variance extracted (AVE), composite
reliability (CR) and Cronbach’s alpha value are reported in Table Hypothesis Path t-value Supported
3. The values of the loadings range from .705 to .923, which are H1 0.251 2.003* Yes
above the recommended value of .70. The values of composite H2 0.016 0.137 No
reliability are above the acceptable value of .70. The AVE value H3 0.360 3.493*** Yes
is above the recommend threshold of .50, thus, the convergent H4 0.270 2.487** Yes
validity is confirmed. H5 0.381 3.958*** Yes
Table 3. Results of reliability and validity analysis H6 0.242 2.356* Yes
Standard Alpha
Factor Item AVEa CR
loading Value
Top TMS1 0.923
Managemen TMS2 0.922 0.848 0.944 0.910
t Support TMS3 0.918
ABC2 0.852
Absorptive
ABC3 0.886 0.750 0.900 0.838
capacity
ABC4 0.860
COM1 0.837
Compati- COM2 0.852
bility 0.731 0.916 0.878
COM3 0.865
(COM)
COM4 0.865
COP1 0.843
Complexity
COP2 0.797 0.759 0.904 0.864
(COP)
COP3 0.966 Figure 3. PLS test of research model (non-significant controls
Adopt1 0.900 omitted to improve readability).
Adoption Adopt2 0.943 0.843 0.942 0.907
Adopt3 0.911 5. DISCUSSION
a
AVE and reliability coefficients are only relevant for reflective This paper developed a research model aimed at understanding
constructs the combined and relative effects of institutional pressures, top
management support, and the absorptive capacity of the firm on
cloud computing adoption. The model was tested within the South
4.3 Hypothesis Testing African context using a survey methodology.
PLS was then used to test the structural model of hypothesized
relationships. The PLS results are reported in Figure 3. The R2 We found support for institutional theory as a theoretical
value for adoption of cloud computing is .607, which means the perspective. Specifically, by confirming hypothesis 1, we show
model explains 60.7% of variance.. Mimetic pressure has a that organisations are modeling their response to cloud computing
significant positive effect on adoption (β=0.251, p<0.01) with the on other organisations they perceived to be legitimate and
two first-order constructs both significant sources of mimetic successful [18, 19]. If there are enough similar organisations that
pressure. Extent of cloud adoption among customers and behave in a certain manner then functioning in that manner
pressures from professional and industry bodies were significant becomes important to confer legitimacy and avoid being
normative pressures whilst extent of adoption by suppliers was identified as less innovative [20]. The social context of South
not. Both mimetic (β=0.270, p<0.05) and normative pressures

35
African firms has important implications for their IT adoption influence of both external and internal factors driving cloud
behaviors. computing adoption.
We also confirm hypotheses 3, 4 and 5. Top management support
has the strongest direct effect on adoption and our results thus
confirm the important role played by top managers in creating a 7. REFERENCES
positive environment for IT innovation adoption [14, 34]. The
finding that top management championship is an important [1] Greengard, S. 2010. Cloud computing and developing
mediating variable in the effects of both mimetic and normative nations. Communications of the ACM 53, 5, 18-20.
institutional pressures also confirms that top managers play an
important boundary spanning role, and are thus the focal point for [2] Misra, S. C., and Mondal, A. 2011. Identification of a
external pressure [14]. Institutional pressures influence company’s suitability for the adoption of cloud computing
organisational action through their effects on top management’s and modelling its corresponding Return on Investment.
beliefs in the relative advantages offered by the technology [25]. Mathematical and Computer Modelling 53, 504-521.
This is particularly the case for normative pressures, which were [3] Rivard, S., Raymond, L., and Verreault, D. 2006. Resource-
fully mediated by top management. This suggests that based view and competitive strategy: An integrated model of
institutional norms regarding the adoption of an IT innovation can the contribution of information technology to firm
influence organizational action but must first be interpreted by top performance. Journal of Strategic Information Systems 15,
managers before they are acted upon [14]. Other factors capable 29-50.
of moderating top managements’ beliefs in the relative advantages [4] Dhar, S. 2012. From outsourcing to Cloud computing:
of an innovation are therefore capable of diluting the effects of evolution of IT services. Management Research Review 35,
normative pressures to conform. The effects of mimetic pressures 8, 664 – 675.
on adoption are only partially mediated by top management and
thus future research should consider whether other internal factors [5] Wilson, P. 2011. Positive perspectives on cloud security.
translate mimetic pressures into adoption behaviour. Information Security Technical Report 16, 97-101.

Results also confirmed hypothesis 6. We show that firms with [6] Marston, S., Li, Z., Bandyopadhyay, S., Zhang, J., and
greater existing knowledge and abilities related to cloud Ghalsasi, A. 2011. Cloud Computing – The Business
computing and who train employees on cloud computing Perspective. Decision Support Systems 51, 176-189.
technologies have a greater absorptive capacity to adopt cloud [7] Babcock, C. 2012. Cloud's Thorniest Question: Does it Pay
computing. We thus confirm that absorptive capacity of South Off? InformationWeek, 4 June, 1-6.
African firms is important to lowering the knowledge barriers [8] Sultan, N. A. 2011. Reaching for the Cloud: How SMEs Can
surrounding new IT and enabling these firms to more rapidly Manage. International Journal of Information Management
adopt [30]. Without developing absorptive capacity, firms 31, 272-278.
constrain their potential and cannot easily recognise and
appreciate the opportunities or threats surrounding new [9] Lin, A. and Chen, N.C. 2012. Cloud computing as an
innovations. innovation: Perception, attitude, and adoption. International
Journal of Information Management 51, 176-189.
It is important to note some limitations of our study. The study
[10] Geczy, P., Izumi, N., and Hasida, K. 2012. Cloudsourcing:
was conducted in South Africa and results may not be fully
Managing Cloud Adoption. Global Journal of Business
generalisable to firms in other countries. Our sample size (87) is
Research 6, 2, 57-70.
also fairly small and non-response bias may further limit the
external validity of our findings. Our study is cross-sectional and [11] Kshetri, N. 2010. Cloud Computing in Developing
therefore causal inferences are only made with respect to theory. Economies. Computer 43, 10, 47-55.
Future studies may wish to adopt longitudinal designs to [12] Burns, C. 2012. Infrastructure as a Service: Overview.
strengthen causal inferences. Future studies may also consider Enterprise Cloud Services, April, 27-29.
other factors e.g. regulatory environment that may be important in
certain industry contexts, as well as other internal organisational [13] Low, C., Chen, Y., and Wu, M. 2011. Understanding the
factors such as the track record of the in-house IS function that Determinants of Cloud Computing Adoption. Industrial
might influence cloud adoption. Management and Data Systems 111, 7, 1006-1023.
[14] Liang, H., Saraf, N., Hu, Q., and Xue, Y. 2007. Assimilation
6. CONCLUSION of Enterprise Systems: The effect of Institutional Pressures
Cloud computing is a technology that holds much promise for and the Mediating Role of Top Management. MIS Quarterly
South African organisations, yet there are a number of concerns 31, 1, 59-87.
and uncertainties surrounding it. This study contributed by [15] Oliveira, T. and Martins, M. F. 2011. Literature Review of
developing and testing a research model aimed at understanding Information Technology Adoption Models at Firm Level.
organisational adoption of cloud computing. Data from 87 South The Electronic Journal Information Systems Evaluation 14,
African firms was used to test hypothesized relationships. Results 1, 110-121.
have confirmed the utility of institutional and absorptive capacity [16] Li, Y.H. 2008. An empirical investigation on the
theories as explanations for cloud computing adoption, and we determinants of e-procurement adoption in Chinese
highlight the relative importance of top management support. manufacturing enterprises. 15th International Conference on
Through our study, we have thus offered useful insights into the Management Science & Engineering, California, USA, Vols
I and II, Conference Proceedings, 32-37.

36
[17] Scott, W.R. 1995. Institutions and Organizations, Sage [30] Roberts, N., Galluch, P.S., Dinger, M., and Grover, V. 2012.
Publications, Thousand Oaks, CA. Absorptive capacity and information systems research:
[18] DiMaggio, P. J., and Powell, W. W. 1983. The Iron Cage Review, synthesis, and directions for future research. MIS
Revisited: Institutional Isomorphism and Collective Quarterly 36, 2, 625-648.
Rationality in Organizational Fields. American Sociological [31] Armstrong, C. P. and Sambamurthy, V. 1999. Information
Review 48, 2, 147-160. Technology Assimilation in Firms: The Influence of Senior
[19] Ravichandran, T., Han, S., and Hasan, I. 2009. Effects of Leaders and IT Infrastructures. Information Systems
Institutional Pressures on Information Technology Research 10, 4, 304-327.
Investments: An Empirical Investigation. IEEE Transactions [32] Elbashir, M.Z., Collier, P.A., and Sutton, S.G. 2011. The role
on Engineering Management 56, 4, 677-691. of organizational absorptive capacity in strategic use of
[20] Teo, H. H., Wei, K. K., and Benbasat, I. 2003. Predicting business intelligence to support integrated management
Intention to Adopt Interorganizational Linkages: An control systems. The Accounting Review 86, 1, 155-184.
Institutional Perspective. MIS Quarterly 27, 1, 19-49. [33] Soares-Aguiar, A. and Palma-Dos-Reis, A. 2008. Why do
[21] Weerakkody, V., Dwivedi, K.Y., and Irani, Z. 2009. The firms adopt e-procurement systems? Using logistic
Diffusion and Use of Institutional Theory: A Cross- regression to empirically test a conceptual model. IEEE
disciplinary longitudinal Literature Survey. Journal of Transactions on Engineering Management 55, 1, 120-133.
Information Technology 24, 354-368. [34] Premkumar, G. 2003. A meta-analysis of research on
[22] Khalifa, M. and Davison, R.M. 2006. SME adoption of IT: information technology implementation in small business.
The case of electronic trading systems. IEEE Transactions on Journal of Organizational Computing an Electronic
Engineering Management 53, 2, 275-284. Commerce 13, 2, 91-121.

[23] Bharati, P., Zhang, C., and Chaudhury, A. 2014. Social [35] Enns, H.G., Huff, S.L., Golden B.R. 2001. How CIOs obtain
media assimilation in firms: Investigating the roles of peer commitment to strategic IS proposals: barriers and
absorptive capacity and institutional pressures. Information facilitators. Journal of Strategic Information Systems 10, 3-
system frontiers 16, 257-272. 14.

[24] Liu, H., Ke, W., Wei, K.K., Gu, J., and Chen, H. 2010. The [36] Jeyaraj, A., Rottman, J.W., Lacity, M.C. 2006. A review of
role of institutional pressures and organizational culture in predictors, linkages, and biases in IT innovation adoption.
the firm’s intention to adopt internet-enabled supply chain Journal of Information Technology 21, 1–23.
management systems. Journal of Operations Management 28, [37] Ye, F., Zhao, X., Prahinski, C., and Li, Y. 2013. The impact
372-384. of institutional pressures, top managers’ posture and perverse
[25] Pahlke, I., Beck, R., and Vykoukal, J. 2011. Follow the pack logistics on performance-Evidence from China. International
or make an independent decision? How environmental Journal of Production Economics 143, 132-143.
turbulence affects ICT sourcing decisions. In Proceedings of [38] Zheng, D., Chen, J., Huang, L., and Zhang, C. 2013. E-
ICIS, Paper 10. government adoption in public administration organizations:
[26] Cooper, B. R., and Zmud, W. R. 1990. Information Integrating institutional theory perspective and resource-
Technology Implementation Research: A Technological based view. European Journal of Information Systems 22,
Diffusion Approach. Management Science 36, 2, 123-139. 221-234.

[27] Cohen, W.M. and Levinthal, D.A. 1990. Absorptive [39] Zahra, S.A., and George, G. 2002. Absorptive Capacity: A
capacity: A new perspective on learning and innovation. review, reconceptualization, and extension. The Academy of
Administrative Science Quarterly 35, 1, 128-152. Management Review 27, 2, 185-203.

[28] Liao, S.H., Wu, C.C., Hu, D.C., and Tsuei, G.A. 2009. [40] Tornatzky, L.G., and Klein, K..J. 1982. Innovation
Knowledge acquisition, absorptive capacity and innovation characteristics and innovation adoption implementation: a
capability: An empirical study of Taiwan’s knowledge- meta-analysis of findings. IEEE Transactions on Engineering
intensive industries. World Academy of Science Engineering Management 29, 1, 28-45.
and Technology, 53, 160-166. [41] Kostova, T., and Roth, K. 2002. Adoption of an
[29] Corrales M. (2010). From Absorptive Capacity to Best organizational practice by subsidiaries of multinational
Practices to Mobilize Knowledge and Open Innovation. corporations: Institutional and relational effects. The
Proceedings of Melbourne 2010 Knowledge Cities World Academy of Management Journal 45, 1, 215-233.
Summit. Melbourne, 16-19 Nov 2010.

37

You might also like