J. Serv. Sci. & Management. 2008, 1: 1-9
Published Online June 2008 in SciRes (www.SRPublishing.org/journal/jssm)
Copyright © 2008 SciRes JSSM
BBA Decision Model for Information Systems Outsourcing:
Using a Multicriteria Method
Jian-Jun WangP
P, Zheng-Kui LinP
P and Hao HuangP
PSchool of Ma n agement, Dalian Univ ersity of Te chnology, Dalian 116024, P.R .China
PSchool of Computer Science & Technology, Dalian Maritime University, Dalian 116026, P.R.China
PSchool of Economic a n d Management, Beihang University, Beijing 100083, P.R.China
An ever-increasing trend in today’s firms is to exploit outsourcing for those information systems (IS) functions deemed
to be outside the company’s core competence. Given the multi-attribute nature of IS outsourcing decision, this paper
argues that five facto rs, including, strategy, economics, risk, environment an d quality, should be con sidered for IS out-
sourcing decisions, and proposes the use of analytic hierarchy process (AHP) and improved ELECTREIII as aids in
making IS outsourcing decisions. The AHP is used to analyze the structure of the outsourcing problem and determine
weights of the criteria, and the improved ELECTREIII method is used for final ranking. It shows by means of an appli-
cation that the hybrid method is very well suited as a decision-making tool fo r the IS outsourcing decision. Finally, po-
tential issues for future research are presented.
Keywords: Information systems outsourcing, Multiple criteria analysis, AHP, ELECTREIII
1. Introduction
Information systems (IS) outsourcing can be defined as “a
significant contribution by external vendors in the physi-
cal and/or human resources associated with the entire or
specific components of the IS infrastructure in the user
organization”[1]. IS outsourcing is a growing phenome-
non in a wide variety of industries. According to the
Gartner Group, overall spending in information technol-
ogy (IT) outsourcing reached177 billion in 2003, and it
is predicted to rise to236 billion in 2007(Oh, 2005).
Yet in many cases, outsourced IS projects have failed. For
example, in one study, IS managers reported only a 33%
satisfaction with outsourced IS services, as compared
with a satisfaction rate of 70%-80% for outsourced
non-IS services [2 ]. Wrong IS outsourcing decision is one
of critical reasons which cause the high IS outsourcing
failure [3]. Therefore, the scientific IS outsourcing deci-
sion process is very important to increase the success rate
of outsourcing. The problem of how to scientifically
make the IS outsourcing decisions tends to be an impor-
tant issue facing organizations in today’s rapidly chang-
ing business environment [4].
Business practitioners recognize that IS outsourcing is
one of the m any t ools in thei r toolki t to desi gn and m anage
their business and potentially has a place in most strategic
plans. Although there is a wealth of academic literature
examining outsourcing, it generally addresses the decision
whether or not to outsource: the go/no go choice. There is
little academic literature that address the IS outsourcing
decision in a quantitative way. This paper will apply a
hybrid of analytic hierarchy process (AHP) and improved
ELECTRE III methods to the “IS outs o urcing decision” in
an effort to demonstrate one quantitative approach to this
complex decision.
The remainder of this paper is organized into five sec-
tions. Section 2 presents a brief literature review of the
existing decision models related to the IS outsourcing
decision. Section 3 will briefly describe the two proposed
methodologies. In section 4 we argue five factors pro-
posed as the p rincipal criteria for outsourcing decision . In
Section 5 a description of proposed methodology is fol-
lowed by an application. In Section 6, we present our
conclusions a nd future rese arch.
2. Literature Review
A number of decision frameworks appear in academic
literature to support the outsourcing decision-making
process. Lacity, Willcocks and Feeny (1996) argued that
deciding the outsourcing of IT activities just by strategic or
commodities was fallacious and senior executives might
mistakenly classify all IS activities as commodities.
Therefore, they presented a 2x2 decision matrix guiding
the selection of outsourcing candidates based on the
business, economic, and technical factors[5].
Venkatraman (1997) viewed information technology as
a portfolio of IS elements that were cost centers, service
centers, investment centers and profit centers (which are
collectively referred to as an IS “value centers”). His focus
on “four i ndep endent source s of val ue from IS resource ” is
useful for i dentifyi ng IS elem ents that might be conside red
for outsourci n g [6] .
2 Jian-Jun Wang, Zheng-Kui Lin and Hao Huang
Copyright © 2008 SciRes JSSM
Yang and Huang (2000) argued that five factors, in-
cluding m anagem en t, strat egy, econom ics, technolo gy a nd
quality, should be considered for outsourcing decision,
then they used the analytic hierarchy process (AHP)
method to help users in structuring the outsourcing prob-
Roy and Aubert (2002) presented an IT outsourcing
decision model based on the resource theory[8]. Hsu et al
(2004) analyzed the IS outsourcing decision problems by
case-based reasoning (CBR) method[9]. Aubert et al
[2004] proposed an IT outsourcing decision model relied
on transaction costs and incomplete contracts theo-
According to the above-mentioned literature, research-
ers put so many outsourcing decision strategies and de-
terminants to practitioners, but current practice remains in
the stage of conceptual discussion as to how to outsource
the IS activities. Furthermore, only the AHP method offers
a quantitative magnitude for judgment among these
strategies and determinants. But AHP approach has the
disadvantage that the number of pairwise comparisons to
be made, may become very large (more specifically:
n(n-1)/2). Another critical disadvantage of the AHP
method is that compensation betwee n good scores on som e
criteria and bad scores on other criteria can occur. Other
strategies or determinants, for instance, decision matrix,
transaction cost or CBR method, are too narrow to help the
practitioners determine if their systems could be out-
sourced or to examine the priorities among ma ny pote ntial
IS outsourcing projects. Ineffectiv e outsourcing activities,
derived from improper strategy or method, would lead to
loss core competencies and capabilities, exposure to un-
expected risk and business failures.
Facing the problem of how to decide the priority of
those IS which have been decided to outsource, we pro-
pose a hybrid multi-criteria decision method for the IS
outsourcing decision making. We shall use the AHP
method to anal yze the structure of the outso urcing problem
and determine the weights of criteria, and use the improved
ELECTRE III method for final ranking. The purpose of
this work is to offer a quantitative d ecision mod el that can
help practitioners set priority and reap the most benefits
from outsourcing.
3. The AHP and Electre III Method
3.1. The AHP Method
The AHP, developed by Saaty (1980)[11], is a technique
for considering data or information about a decision in a
systematic manner [12]. The AHP mainly addresses how
to solve decision problems with uncertainty and with
multiple criteria characteristics. It is based on three prin-
ciples: first, constructing the hierarchy; second, priority
setting, and third, logical con sistency.
3.1.1. Construction the hierarchy
A complex decision problem, centered round measuring
contributions to an over objective or focus, is structured
and decomposed in to sub-problems (sub-objectives, crite-
ria, alternatives, etc), within hierarchy.
3.1.2. Priori ty setting
The relative “priority” given to each element in the hier-
archy is determined by comparing pairwise the contribu-
tion of each element at a lower level in terms of the crite-
ria (or elements) with a causal relationship exists. In AHP
multiple paired comparisons are based on a standardized
comparison scale of nine levels (see Table 1, Saaty 1980).
Table 1. Scale of Relative Importance
Intensity of impor-
tance Definition
1 Equal importance
2 Weak
3 Moderate importance
4 Moderate plus
5 Strong importance
6 Strong plus
7 Very strong or demon-
strated importance
8 Very, very strong
9 Extreme importance
Let 1, 2,,{}
nCC==Lbe the set of criteria. The re-
sult of the pairwise comparison on n criteria can be sum-
marized in a ()nn
evaluation matrix
in which every
element ij
ais the quotient of weights of the criteria, as
shown in (1).
(),(, 1,,)
=L (1)
The relative priorities are given by the right eigenvec-
tor (w) corresponding to the largest eigenvector (max
as shown in (2).
Aw w
In case the pairwise comparisons are completely con-
sistent, the matrix
has rank 1 andmax n
=. In that
case, weights can be obtained by normalizing any of the
rows or columns of matrix
The procedure described above is repeated for all sub-
systems in the hierarchy. In order to synthesize the vari-
ous priority vectors, these vectors are weighted with the
global priority of the parent criteria and synthesized. This
process starts at the top of the hierarchy. As a result, the
overall relative priority to be given to the lowest level
elements is obtained. These overall, relative priorities
indicate the degree to which the alternatives contribute to
the focus. These priorities represent a synthesis of the
local priorities, and reflect an evaluation process that
permits to integrate the perspectives of the various stake-
holders involved [13].
A Decision Model for Information Systems Outsourcing: Using a Multicriteria Method 3
Copyright © 2008 SciRes JSSM
3.1.3. Consi s tency check
A measure of consistency of the given pairwise compari-
son is needed. The consistency is defined by the relation
between the entries of A:ij jkik
aa a
=. The “consis-
tency index” (CI) is given by (3).
CI() / (1)nn
=− −
The final consistency ratio (CR), on the basis of which
one can conclude whether the evaluations are sufficiently
consistent, is calculated as the ratio of the consistency
index (CI) and the random consistency index (RI), as in-
dicated in (4). The number 0.1 is the accepted upper limit
for CR. If the final consistency ratio exceeds the number,
the evaluation procedure has to be repeated to improve
consistency. The measurement of consistency can be used
to evaluate the consistency of decision makers as well as
the consistency of all the hierarchy.
CR=CI/RI (4)
3.2. The ELECTRE III Method
3.2.1. The normal ELECTRE III method
ELECTREIII method is a non-compensatory, MCDM
technique. It uses various mathematical functions to indi-
cate the degree of dominance of one alternative or group
of alternatives over the remaining ones. It also facilitates
comparisons between alternative schemes by assigning
weights to decision criteria. The outranking relationships
between alternatives are constructed and exploited even-
A discrete multiple criteria decision making problem is
usually formulated by a set of alternatives
xx x=L, a
set of criteria
Ccc c=Land a set of func-
Ggg g=L. The real-valued functions de-
fined on the setXso that()
xrepresents the perform-
ance of the alternativej
on the criterionl
c. Without loss
of generality we assume that all the objective functions
are to be maximized.
A pseudo-criterion is a preference model including
three different thresholds: a preference threshold
pgx , an indifference threshold (())
qgx and a
veto threshold(( ))
vgx for each criterion
Ccc c=L. These thresholds may be constants, lin-
ear or affine functions of()
x in the form [14].
(()) ()
lljpl pllj
pgx gx
and ,,
(()) ()
lljql qllj
qgx gx
For every criterionl
c, the preference and indifference
threshold model is as follows. i
is preferred toj
lil jl l j
is weakly preferred toj
() (())()()(())
lj lljliljllj
x qgxgx gxpgx
≤+p, and i
indifferent toj
() (())()
lj lljli
xqgx gx
() (())()
lil lil j
x qgxgx
Where (( ))
pgx and (())
qgx are preference
and indifference thresholds, respectively and
(()) (())0
llj llj
pgx qgxff
. Weak preference is sup posed to
describe the decision maker’s hesitation between indif-
ference and preference.
In ELECTRE III method, one considers an outranking
degree (,)
Sx xdescribing the outranking credibility
of i
over j
taking its values between 0 and 1. The value
of (,)
Sx xis defined based on so-called concordance
and discordance indices. A concordance index
(, )
Cx xis computed for each pair of alterna-
tives(, )
(, )(, )
ij llij
Cx xwc x x
= (5)
Where (1,,)
wl m
Lis the weight of each criterion,
() ()()
()() ()()
xgxgxgp xgxgpxg
xxc (7)
A discordance index(, )
dxxis defined for each crite-
rion l
() ()()
() ()()
()() ()()
xxd (8)
pgx is the preference threshold value and
vgx is the veto threshold value of each l
and (()) (())
vgx pgxf.
4 Jian-Jun Wang, Zheng-Kui Lin and Hao Huang
Copyright © 2008 SciRes JSSM
Finally, the degree of outranking is defined by
() ()
jixxJl ji
ji xxJ
where (, )
xx is the set of criteria for which
(, )(, )
li ji j
dxx Cxxf. The complete set of outranking
degree is assembled as shown in the following credibility
11 121
21 222
(,)(, )(, )
(,) (,)(,)
(,) (,)(,)
nn nn
Sx xSx xSx x
Sx xSx xSx x
Sx xSx xSx x
S (10)
The ranking of the decision alternatives in ELECTRE
III is carried out by a distillation procedure, where the
alternatives are ranked based on their qualification from
the best to the worst (descending distillation process) and
from the worst to the best (ascending distillation proc-
ess).The final partial order of the alternatives is built
based on these two complete orders. The descending dis-
tillation process is as follows.
Let ,
max( ,)
ij ij
xx X
Sx x
=. Determine a “credibility
value” such that only values of (, )
Sx xthat are suffi-
ciently close to
are considered; that is, ()
. Thus
if 1
=, let()0.15s
=. Define the matrixTas:
1if (,)()
(, )0otherwise
Sx xs
T (11)
Further, define the qualification of each alternative
−− )(i
xQ as the number of alternatives that are out-
ranked byi
minus the number of alternatives which out-
rank i
Qx is simply the row minus the column sum of
the matrixT. The set of alternatives having the largest
qualification is the first distillate of1
. If 1
only one alternative, repeat the previous procedure
with 1
.Otherwise, apply the same procedure in-
side 1
. If distillate2
contains only one alternative, the
procedure is started in12
D(unless the set is empty);
otherwise it is applied within2
, and so on until 1
used up. The procedure is then repeated starting
with 1
. The outcome is the first preorder.
The ascending distillation is carried out in a similar
fashion except that alternatives with the smallest (rather
than the largest) qualification are retained first.
3.2.2. An improved ranking method for ELECTRE III
The normal ranking of ELECTRE III requires an addi-
tional threshold to be introduced. The weakness of this
handing is that the ranking of the alternatives depends on
the size of this threshold for which there exists no “cor-
rect” value. Additionally, the final ranking is not com-
Aiming at the ranking problems in ELECTRE III, we
present a new ranking method by introducing three defi-
nitions concordance credibility degree, discordance
credibility degree and net credibility degree, into ELEC-
TRE III method.
The concordance credibility degree is defined by
()(, ),
Sx xxX
The concordance credibility degree is the measure of
the outranking character ofi
(how i
dominates all the
other alternatives of
The discordance credibility degree is defined by
() (,),
Sx xxX
The discordance credibility degree gives the outranked
character of i
(how i
is dominated by all the other al-
ternatives ofX)
The net credibility degree is defined by
()() (),
The net credibility degree represents a value function,
where a higher value reflects a higher attractiveness of
alternative i
Final ranking
All the alternatives can be completely ranked by the net
credibility degree.
An application
This application is based on a real-life case study where
the ELECTRE III method was use to help choose route
for Dublin port motorway [15].
The first credibility matrix in [15] is as follows:
A Decision Model for Information Systems Outsourcing: Using a Multicriteria Method 5
Copyright © 2008 SciRes JSSM
B2/B30.625 0.750.8750.8750.875 0.875
B40.8751110.875 0.875
B60.8751110.875 0.875
B5/7it0.8750.875 0.8750.8750.8750.875
B5/7bt 0.8751110.875 0.875
B5/7ht0.750.875 0.8750.8750.87
B80.688 00000
The final ranking based on the normal method is as
follows [15]:
{B4,B6,B5/7bt} {B2/B3,B5/7hb} {B5/7it,B8}→→
With 1
and (12)-(14) leads to the final values of con-
cordance credibility degree, discordance credibility de-
gree and net credibility degree of alternatives in Table 2.
Table 2. Degrees of alternatives (1
B2/B3 4.875 4.938 063.0
B4 5.625 4.375 1.25
B6 5.625 4.5 1.125
B5/7it 5.25 4.75 0.5
B5/7bt 5.625 4.625 1
B5/7ht 5.25 4.375 0.875
B8 0.688 5.375 687.4
The final ranking based on the new method is as fol-
B4 B6B5/7bt B5/7htB5/7it
B5/7itB2/B3 B8
→→ → →
The second credibility matrix in [15] is as follows:
B2/B3B4B6B5/7it B5/7bt B5/7htB8
B2/B30.45 0.610.810.81
B40.931110.93 0.93
B60.931110.93 0.93
B5/7bt 0.931110.93 0.93
B5/7ht0.750.82 0.820.820.821
B80.69 00000
The final ranking based on the normal method is as
follows [15]:
{B4,B6,B5/7it,B5/7bt} B5/7htB2/B3 B8→→→
and (12)-(14) leads to the final values of con-
cordance credibility degree, discordance credibility de-
gree and net credibility degree of alternatives in Table 3.
The final ranking based on the new method is as fol-
B4B6B5/7bt B5/7it B5/7ht
B5/7htB2/B3 B8
→→→ →
From above comparison, the final r anking based on the
new method is complete ranking, while the normal is par-
tial ranking.
Table 3. Degrees of alternatives (2
B2/B3 4.29 5.16 87.0
B4 5.79 4.24 1.55
B6 5.79 4.4 1.39
B5/7it 5.7 4.63 1.07
B5/7bt 5.79 4.6 1.19
B5/7ht 5.03 4.53 0.5
B8 0.69 5.52 83.4
The normal ranking method of ELECTRE III suffers
from the complicated ranking process, which requires an
additional threshold to be introduced. The weakness of
this handing is that the ranking of the alternatives depends
on the size of this threshold for which there exists no
“correct” value. Additionally, the normal ranking method
also suffers from incomplete ranking result. Aiming at the
ranking problems in ELECTRE III, the present research
develops a new ranking method. Compare to the normal
ranking method, the new ranking method is simple and
ranking result is complete.
4. Performance Criteria
There have been a lot of attempts to find out all factors of
outsourcing decision, but the problem has not been theo-
retically solved. The choice of factors has be en selected in
agreement with a group of experts and managers. Another
group might have selected a somewhat different set of
factors. Firms should select all factors which can affect
organizations benefit as possible as they can. A careful
examination of factors used before concludes that five
dimensions, strategy, economics, risk, environment and
quality, should be included.
4.1. Strategy
TFor strategy, firms need to focus on their core activities
and outsource noncore activities. IS outsourcing allows
management to focus available IS talent on important and
strategic IT applications rather than the mundane and rou-
tine activities. The internal operations and outsourced
operations should then work in unions striving to opti-
mize flexibility and responsiveness to customer and in-
ternal needs, and minimize unnecessary paperwork and
bureaucracy. In addition, the firms can make strategic
alliance with vendors to make up the shortage of re-
sources; resources include new technologies and profes-
sional workers. From strategic alliances, the firm even
can develop and market new products. Other strategic
consideration includes sharing risks and accelerating the
time of product to market [5][7][16][17][18][ 19]T.
6 Jian-Jun Wang, Zheng-Kui Lin and Hao Huang
Copyright © 2008 SciRes JSSM
4.2. Economics
For economics, the major consideration of a firm is to
reduce costs of information systems. Because the vendors
have a better management skill as well as higher produc-
tivity per employee, the costs can be reduced. Meanwhile,
Because of the scale of economics vendors have invested
in the hardware, software and human resources, the cost
can be reduced. Another consideration of economics is
financial flexibility. Because of outsourcing, the facilities
and employee would be transferred to the vendor side,
which transform fixed costs into variable costs, resulting
in increasing financial flexibility [7][9][20][21].
4.3. Risk
For risk, it is rare to experience opportunities in organiza-
tional life where the managerial actions taken to produce
benefits are not associated with potential risks either. This
is most certainly the case with IS outsourcing. The most
prominent risks in outsourcing are information security
concerns an d loss of management co ntrol. Khalfan [2004]
noted theses two factors were coupled with hidden costs
in outsourcing [22].
If a labor union exists, a firm should first explore its
negative effect before deciding to outsource, since out-
sourcing is accompanied with some possibility of layoffs.
Companies often have to deal with low employee morale
as a result of outsourcing, and low employee morale in
turn affects productivity. It has been noted that often a
large proportion of IT staff are laid off as a consequence
of an outsourcing contract. This can cause a lot of distur-
bance in the client company.
Other risks that have to be dealt with include: loss of
core competence, loss of internal technical knowledge,
loss of flexibility, damaging the firm’s innovative capa-
bility, increasing information services management com-
plexity, etc [23][24]. As being the factors with benefits,
these risks factors should not be ignored in outsourcing
activities [25] [26].
4.4. Environment
Quinn and Hilmer [1994] indicated that environment fac-
tors such as market maturity, market depth, and the num-
ber of suppliers influences the level of outsourcing [16].
There are times when contestability explains market ma-
turity; a contestable market means that while only a few
firms can immediately provide the service now, many
other firms are intending to provide the service if the
price paid by the firm exceeds the average cost of vendors.
In addition, the decision to outsource may be induced by
imitative behavior among firms [27]. For example, Ko-
dak’s outsourcing decision made many other firms begin
to consider IS outs o urci n g as a viable alternative [28].
4.5. Quality
TFor quality, because vendors may have access to more
technological environments, have more qualified or more
motivated personnel, provide a greater breadth of services,
and simply be more committed than internal staff to mak-
ing the alliance with the customer work well, outsourcing
can improve the quality and services of the internal IS
department. Therefore, good quality of service and good
relationship are the significant success factors of out-
sourcing [7] [25[29].
5. An Application
Based on IS outsourcing decision problem presented in
Section 1, an example is used to illustrate how the com-
bined AHP and improved ELECTREIII model support
decision maker on the IS outsourcing decision making.
5.1. The Problem Faced
A bookstore wants to outsource parts of IT functions, they
think about the management and cost issues and want to
know how to decide which systems should be outsourced
The candidate systems for outsourcing are facilities
management (P1), development of internet homepage
(P2), maintenance of the customer relationship manage-
ment information system (P3), development of the sup-
plier relationship management information system (P4),
development and maintenance of the online transaction
processing system (P5).
The leader of the task force is the vice president, while
members include IS department manager, a senior engi-
neer, business department manager, finance department
manager, planning department manager and five profes-
sional consultants. The vice president convened a meeting
to discuss this problem. After some discussion, they em-
ploy the hybrid of AHP and ELECTRE III methods in the
decision process.
5.2. Structure of the Problem
After some debate, according to the AHP method, the
Figure 1. The Hierarchy Structure of the
candidate 1Outsourcing
candidate 2 Outsourcing
candidate 5
Select the
outsourcing system
Strategy Environment Economics Risk
A Decision Model for Information Systems Outsourcing: Using a Multicriteria Method 7
Copyright © 2008 SciRes JSSM
task force depicts a hierarchy structure as shown in Figure
5.3. Determination of the Weights
Following the computing method described in the AHP,
experts began to compare the factors of the structure. Af-
ter that, they got the square matrix as shown in Table 4.
Table 4. The Square Matrix.
(C1- Strategy; C2- Economics; C3- Environment;C4-
Risk;C5- Qu a lity)
C1 C2 C3 C4 C5
C1 1 3 4 2 4
C2 1/3 1 1 1/3 1
C3 1/4 1 1 1/3 1
C4 1/2 3 3 1 3
C5 1/4 1 1 1/3 1
According to (1)-(4), we got
max 5.0394
5.4. Evaluation of Alternatives
All outsourcing candidates were evaluated by experts.
According to the criteria of strategy, environment, risk
and quality, a qualitative impact value is used, expressed
on a qualitative scale (judgment on a series of ordered
semantic values; each semantic value included in the set
{very weak, weak, common, good, very good} is associ-
ated with a numerical value {1, 3, 5, 7, 9}, that is used for
the calculations.). The economics indicator is evaluated
by the following formula:
Saving costs / Costs of in house development and
maintenance (%).A55×matrix was produced, as shown in
table 5.
Table 5. Evaluation Matrix (C1- Strategy; C2-
Economics; C 3- Environment;C 4- Ri sk ;C 5- Quality)
Criteria C1 C2 C3 C4 C5
Max/Min Max Max Max Min Max
Weight 0.41 0.11 0.10 0.28 0.10
P1 9 15 9 1 9
P2 7 12 7 3 7
P3 3 8 3 7 3
P4 7 15 5 5 7
P5 7 10 7 7 5
5.5. Final Ranking
Before using the ELECTRE III method to calculate the
indices, for each criterion’s thresholds is defined (see
table 6).
Table 6. The Thresholds of Each Criterion
q p v
Strategy 2 3 8
Economics 2 4 8
Environment 2 3 4
Risk 2 3 7
Quality 2 3 6
According to (5)-(9), we get the outranking degree
0.9411 0.941
According to (12)-(14), we got the values of leaving,
entering and net flows and the complete ranking of alter-
natives in table 7 and in figure 2.
The priorities for outsourcing the five IS are in the fol-
lowing order: facilities management (P1), development of
internet homepage (P2), development of the supplier rela-
tionship management information system (P4), develop-
ment and maintenance of the online transaction process-
ing system (P5), maintenance of the customer relationship
management information system (P3). If the bookstore
want to outsource two IS activities first, we know that the
facilities management and development of internet
homepage would be outsourced.
Table 7. Values of Leaving, Entering and Net Flows
P1 4 1.20 2.80
P2 3.88 2.72 1.16
P3 0.15 4 -3.85
P4 3 2.98 -0.02
P5 2.87 3 -0.13
6. Conclusions and Future Researches
IS outsourcing is emerging as a flexible and powerful
P1 P2 P4 P5 P3
Figure 2. Final Ranking
8 Jian-Jun Wang, Zheng-Kui Lin and Hao Huang
Copyright © 2008 SciRes JSSM
management approach chosen by managers to achieve a
wide range of tactical and strategic goals. Outsourcing
firms benefit from cost savings, strategic fitness, im-
proved management effectiveness, technology upgrade,
and the service quality of IS. Moreover, one needs an
operational decision model that can offer systematic steps
and quantitative results to increase the precision of deci-
This study suggests a decision model for IS outsourc-
ing adoption for management, and shows how it may be
applied in a real decision process for IS outsourcing. This
research also argues that firms need to consider more di-
mensions, including economics, strategy, risk, environ-
ment and quality factors. Meanwhile, we offer a decision
model, which developed by AHP and improved ELEC-
TRE III methods, to help the practitioners make better
decisions. Our approach allows to deal with IS outsourc-
ing project selection involving several conflicting per-
formance criteria (qualitative as well quantitative). The
proposed decision model can help practitioners analyze
factors and attributes easily. Because it is a quantitative
process, the practitioners can make better decisions and
obtain better results from outsourcing.
While it has successfully developed the decision model
for IS outsourcing, and with most research efforts, this
study is not without limitations. First, the case study just
based on a small firm. Second, the AHP and ELECTRE
III methods also have their own limitations; such as, in
order to get more professional results, the use of more
advance form of AHP method would be desirable. In ad-
dition, the determinants in the decision model are not
complete. Further studies need to include additional pos-
sible factors through a more extensive literature review
and empirically investigation.
7. Acknowledgements
This work is partially supported by a grant from National
Science Fund for Distinguishe d Young Scholars of Ch ina
(No. 70725004).
The authors also thank the anonymous referees for im-
proving the quality of the paper with their precious and
careful remarks.
[1] Loh, L., & Venkatraman, N, “Determinants of in-
formation technology outsourcing: a cross-sectional
analysis”, Journal of Management Information Sys-
tems, 1992, 9(1), 7-24.
[2] King, W.R, “Developing a sourcing strategy for IS:
A behavioral decision process and framework”,
IEEE Transaction Engineering Management, 2001,
48(1), 15-24
[3] Wu, F., & Li, H.Z, “The effect of knowledge man-
agement on the successful outsourcing of IT/IS”,
Science Research M a na gement, 2004, 25(2), 82-87.
[4] Lee, Jae-N, “The impact of knowledge sharing, or-
ganizational capability and partnership quality on IS
outsourcing success”, Information & Management,
2001, 38, 323-335.
[5] Lacity, M.C., Willcocks, L.P., & Feeny, D.F, “The
value of IT sourcing”, Sloan Management Review,
1996, 37(3), 13-25.
[6] Venkatraman, N, “Beyond outsourcing: Manage-
ment IT resources as a value center”, Sloan Man-
agement Review, 1997, 38(3), 51-64.
[7] Yang, C., & Huang, J, “A decision model for IS
outsourcing”, International Journal of Information
Management, 2000, 20(3), 225-239.
[8] Roy, V., & Aubert, B, “A resources-based analysis
of IT sourcing”, ACM SIGMIS Database, 2002,
33(2), 29-40.
[9] Hsu, C.I., Chiu. C., & Hsu, P.L, “Predicting infor-
mation systems outsourcing success using a hierar-
chical design of case-based reasoning”, Expert Sys-
tems with Application, 2004, 26(3), 435-441.
[10] Aubert, B.A., River d, S., & Patry, M, “A transaction
cost model of IT outsourcing”, Information & Man-
agement, 2004, 41, 921-932.
[11] Saaty, T. L, The analytic hierarchy process. New
York: McGraw Hill, 1980.
[12] Schniederjans M J, Garvin T. “Using the analytic
hierarchy process and multi-objective programming
for the selection of cost drivers in activity-based
costing,” European Journal of Operational Re-
search, 1997, 100: 72 - 80 .
[13] Macharis C, Springael J, De Brucher K, Verbeke A.
“PROMETHEE and AHP: the design of operational
synergies in multicriteria analysis. Strengthening
PROMETHEE with ideas of AHP,” European
Journal of Operational Research, 2004, 153(2):
[14] K. Miettinen, P. Salminen, “Decision-aid for dis-
crete multiple criteria decision making problems
with imprecise data”, European Journal of Opera-
tional Research, 1999, 119: 50-60.
[15] M. Rogers, M. Bruen, “Using ELECTRE III to
choose route for Dublin fort motorway”, Journal of
Transportation Engineering, 2000, 126(4):
[16] Quinn, J.B., & Hilmer, F.B, „Strategic outsourcing”,
Sloan Management Review 1994, 35(4), 43-55.
[17] Quinn, J.B, “Strategic outsourcing”, Sloan Man-
agement Review, 1999, 40(4), 23-26.
[18] Gilley, K.M., & Rasheed, A, “Making more by do-
ing less: Analysis of outsourcing and its effects on
firm performance”, Journal of Management, 2000,
26, 763-790.
[19] Avgerou, C., & Cornford, T, Developing Informa-
tion Systems: Concepts, Issues and Practice (2nd
edn). Basingstoke: Macmillan; 1998.
A Decision Model for Information Systems Outsourcing: Using a Multicriteria Method 9
Copyright © 2008 SciRes JSSM
[20] McFarlan, F.W., & Nolan, R.L, “H ow to manage an
IT outsourcing alliance”, Sloan Management Review,
1995, 36(2), 9-23.
[21] Collins, J., Millen. R., & Beamish, P, “Information
systems outsourcing by large American industrial
firms: Choice and impact”, Information Resources
Management Journal, 199 5, 8( 1) , 5-13.
[22] Khalfan, A.M, “Information security considerations
in IS/IT outsourcing projects: a descriptive case
study of two sectors” International Journal of In-
formation Management, 2004, 24, 29-42.
[23] Bahli, B., Rivard, S, “Validating measures of infor-
mation technology outsourcing risk factors”, Omega,
2005, 33(2): 175-187.
[24] Osei-Bryson, K. Muata., Ngwenyama, O. K, “Man-
aging risks in information systems outsourcing: An
approach to analyzing outsourcing risks and struc-
turing incentive contracts”, European Journal of
Operational Research, 2006, 174(1): 245-264.
[25] Clark, T., Zmud, R., & McCray, G, “The outsourc-
ing of information services: transforming the nature
of business in the information industry”, Journal of
Information Technology, 1995, 10(4), 221-237.
[26] Earl, M.J, “The risks of outsourcing IT”, Sloan
Management Review, 1996, 37(3), 26-32.
[27] Lacity, M.C., Herschheim,R., & Willcocks, L, “Re-
alizing outsourcing expectations”, Information Sys-
tems Management, 1994, 4, 7-18.
[28] Smith, M., Mitra, S., & Narasimhan, S, “Information
systems outsourcing: a study of pre-event firm char-
acteristics”, Journal of Management Information
Systems, 1998, 15, 61-93.
[29] Grover, V., Joong, M., Cheon., & Teng, J.T, “The effect
of service qualify and partnership on the outsourcing of
information systems functions”, Journal of Management
Information Systems, 1996, 12(4), 89-116.
Jian-Jun Wang completed his Ph.D. in Information Systems from Dalian University of Technology, P.R.China. He
has been Assistant Professor at Dalian University of Technology. His main research fields are IS/IT outsourcing, ser-
vice management and e-commerce. He is involved in many national research projects and he has published many re-
search papers in refereed academic journals such as Computers & Operations Research, Chaos, Solitons and Fractals,
International Journal of Pure and Applied Mathematics, International Journal of Knowledge and Systems Sciences, etc.
He served as a reviewer for a number of international refereed academic journals and refereed conferences. Email:
Zheng-Kui Lin completed his Ph.D. in Management Science and Engineering from Dalian University of Technol-
ogy, P.R.China. He has been Associate Professor at Dalian Maritime University. His main research fields are informa-
tion systems and e-commerce. He is author of more than 10 articles published on national and international refereed
journals. Email: dalianjx@163.com
Hao Huang is a Ph.D. candidate in School of Economic and Management at Beihang University, P.R.China. He
works in the field of information management. Email: huang_hao98@yahoo.com.cn
10 Jian-Jun Wang, Zheng-Kui Lin and Hao Huang
Copyright © 2008 SciRes JSSM