International Journal of Computer Applications (0975 – 8887)

Volume 176 – No.7, October 2017

ERP selection: A step-by-step application of AHP

Method
Noureddine Motaki Oualed Kamach
Laboratory LTI Laboratory LTI
ENSA of Tangier ENSA of Tangier
Tangier, Morocco Tangier, Morocco

ABSTRACT Several methods have been applied to ERP selection

A successful implementation of an ERP project in an including scoring, ranking, mathematical optimization, and
organization requires the selection of a suitable ERP system. multi-criteria decision analysis.
Indeed, the selection of an ERP system has a significant effect The analytic hierarchy process (AHP) method is applied for
on future operations and profitability of the enterprise. dealing with the ambiguities involved in the assessment of
Many research works identified major factors influencing the ERP alternatives and relative importance weightings of
selection of an ERP in SMEs, and proposed criteria and criteria. To apply AHP, weights of the criteria are determined
methods related to the optimization of the selection process. after the consultation session realized by the selection team.
To determine weights of the criteria in the AHP hierarchy,
Several ERPs selecting methods use AHP, on one hand in each criterion is compared using a pairwise comparison
order to determine the weight of the criteria, and on the other method with respect to their immediate higher-level criteria in
hand to evaluate ERP systems. Some methods use only AHP, the hierarchy. According to Saaty, quantitative or qualitative
where other methods use a combination of AHP with other assessments can be used in the comparisons: a nine-point
Multi decision criteria Methods (hybrid methods). numerical scale, is recommended for the comparisons.
The first objective of this paper is to present a review of In this paper a step-by-step application of AHP Method will
literature on ERP selection process: after describing the be demonstrated.
methods used to select the ERP systems, the criteria list that
companies use to select their ERP systems is presented. 2. STATE OF THE ART
In literature, several evaluation models have been proposed to
The second objective is to propose a detailed example of improve the selection process of an ERP. A part of these
using the AHP method in an ERP selection context, a step-by- methods has a theoretical foundation based in mathematic
step application of AHP will be demonstrated. programming and decision theories: multi-criteria decision-
making methods (MCDM) are used to prioritize alternatives
Keywords and calculate the relative efficiencies of ERP solutions. Other
ERP, Selection, criteria, Process, AHP
methods have focused on identifying success factors in ERP
1. INTRODUCTION selection process [2].
Companies need to increase productivity, shorten lead times, The review of the state of the art on ERP selection Methods
reduce stock, maximize return on investment, reduce total revealed the following methods: AHP (Analytic Hierarchy
costs, and have a better communication with customers and Process), ANP (Analytic Network Process) [3], [4], [5], [6],
suppliers. [7] PROMETHEE [8], SHERPA (Systematic Help ERP
Acquisition), FL (Fuzzy Logic) [9], PM (Priority Matrix),
Enterprise Resource Planning (ERP) is very important to help TOPSIS (Technique for Order Preference by Similarity to
companies achieve these goals because of its ability to Ideal Solution) [6].
integrate the flow of information, material, and finance, and to
support organizational strategies. An ERP system integrates 2.1 AHP Method
all necessary business functions, such as purchasing, sales, Introduced by Saaty (1980), The AHP method directs how to
product planning, financial, inventory control, through determine the priority of a set of alternatives and the relative
modules that share a single database. However, given the importance of attributes in a multiple criteria decision-making
diversity of the business processes and the sector of activity of problem,
each company, there is no ERP commercial packages can This method has been widely discussed in various studies:
meet all company functionalities or all special business The ERP selection method proposed by Wei, Chien and Wang
requirements. Thus, it is necessary to select a suitable ERP is based on AHP method and presented in seven steps [10].
system that can meet the needs of the company and its The AHP and ANP methods are discussed in the literature,
organizational requirements. and some applications of these methods are described in
Recent studies show that the majority of ERP projects do not research papers. [11], [12].
end in success. One of the most common reasons that can lead Ünal and Güner proposed a methodology based on AHP for
to this failure is the selection phase study [1]. The selection of ERP provider selection for an organization in the textile
a particular ERP system has a decisive effect on the future industry [13]. In the same sector, Cebeci presented another
operations and profitability of the enterprise, this is why the ERP supplier selection methodology based on fuzzy AHP
applied selection criteria for ERP systems and the method [14]. An illustrative example of the application of fuzzy AHP
used should be analyzed very carefully.

15
 International Journal of Computer Applications (0975 – 8887)
Volume 176 – No.7, October 2017

was also carried out for the selection of ERP outsourcing 2.2.1.2 The ERP system must incorporate best
alternatives [15]. practices:
Numerous studies are successfully combined more than one
The ERP reflects the vendor's interpretation of the most
multi-criteria decision-making methods (hybrid efficient way to perform each enterprise business process; the
methodology): A research paper describes a decision-making ERP modules must to be based on the management best
model using a combination of quality function deployment practices, for business process and for industry areas.
(QFD), fuzzy linear regression and AHP [16]. Kilic, Zaim,
2.2.2 Technical criteria:
and Delen proposed another hybrid methodology; they used
fuzzy AHP and TOPSIS for the ERP selection for an airline 2.2.2.1 New trends in the IT industry:
company. [4] The ERP compatibility with the latest trends in IT.
Burak Efe used an integration of fuzzy AHP and fuzzy
TOPSIS (technique for order preference by similarity to ideal 2.2.2.2 Ease of integration and compatibility
solution), an application case is performed on the ERP with other systems:
selection software of an electronic firm. In this application, The compatibility and the integration with other systems is a
the priority values of criteria in ERP selection have been determining criterion for the selection of the ERP. Indeed,
determined by using fuzzy extension of AHP method. The most of the organizations are using independent and
result of the fuzzy TOPSIS model have been used to define indispensable software like CRM, SRM. the ERP system must
the most appropriate alternative with regard to the firm goals. be able to connect and communicate with these products.
[17] From this perspective, compatibility and integration with
Maria Manuela Cruz-Cunha presented an AHP model for the other systems are considered as a crucial criterion for
selection of ERP system. The model’s set of criteria was selecting the ERP solution.
extracted from the literature review and validated by
Portuguese organizations. They applied this model using a 2.2.2.3 Modules independency:
software that eases the application of the AHP process to the The ERP system should be module-independent, where
selection of ERP packages [18]. organizations are free to choose the modules they need only.

2.2 ERP Selection Criteria 2.2.2.4 The ERP ability to integrate different
The life cycle of the ERP system consists of three main platforms and data:
phases: the selection phase (selection of the product, vendor This criterion evaluates the complexity of data migration from
and consultant), the implementation phase (functional and an old information system to the newly installed ERP, and
technical implementation), and the phase of use (from the Go evaluate also the transition from classical office programs to
live step). The selection process plays an important role in the ERP system.
success or failure of the ERP project: a wrong selection of the
ERP system would either fail the project or critically reduce 2.2.2.5 System stability:
the system and hinder the performance of the organization. The tendency of the system to be more stable.
The ERP research papers have suggested several ERP
selection criteria. As presented in the literature, the selection
criteria for an ERP can be organized around the following
2.2.2.6 Flexibility:
Flexibility of the ERP with the future needs of the company
categories [19], [4], [20], [21], [22], [1]:
and the possible modifications of the business processes.
 Functional criteria
 Technical criteria
2.2.3 The criteria for the vendor (ERP Provider):
Vendor position in the ERP market and its experience in the
 The criteria for the vendor ERP systems, Brand image, References…
 The criteria for services provided
 Partners Criteria 2.2.4 The Criteria for services provided:
 Financial criteria
2.2.4.1 Maintainability and support from
 Implementation phases criteria
vendors:
2.2.1 Functional criteria: The majority of organizations require maintainability and
support from suppliers to face technical problems, security
2.2.1.1 ERP compatibility with the enterprise issues and integration difficulties with the application during
business processes: the whole implementation process and after the go live phase.
It is the most important evaluation factor in the ERP selection
process, in order to validate this phase, ERP systems must 2.2.4.2 Training:
satisfy the following conditions: A training program for users is essential before the go live
phase.
 The ERP systems should have adequate or even
more modules related to the organization main 2.2.5 Partner Criteria:
activities. 2.2.5.1 Consultant’s suggestions:
 The ERP systems should support the critical The expertise of consultants in ERP implementation projects
business processes. and their recommendations can be considered among the most
important factors in the ERP selection process.

16
 International Journal of Computer Applications (0975 – 8887)
Volume 176 – No.7, October 2017

ERP systems used by partners (customers and the following table shows the decomposition of each criterion
suppliers): into sub-criteria.
The tendency to have similar or compatible ERPs is Table 1. Sub-criteria
important: EDIs are often used to ensure interconnection
between partner ERPs.
Criterion Sub-criteria
2.2.5.2 Benchmarking:
Adaptability  Compatibility with the enterprise
The level of the ERP use by competing enterprises or
business processes.
enterprises whose business sector is the same.
 Technical constraints.
 System features
2.2.6 Financial criteria:
 Ability to integrate company platforms
This criterion contains the cost types of ERP implementation
and data.
project, including product-licensing cost, ERP implementation
cost, support service and training costs, organizational change Financial  Service & support cost.
management cost (OCM).
 Product License.
 Implementation cost
2.2.7 Implementation phases criteria:  Budget of the company
2.2.7.1 Complexity of using ERP:
This criterion evaluates the complexity of understanding and Simplicity  Ease of use
use of ERP by end users, key users and managers. ERPs can  Ergonomic software
be distinguished according to the usability of the user  Complexity system
interfaces and the ergonomic rules applied.
Provider services  Maintainability from provider
2.2.7.2 The duration and complexity of ERP  Support from provider
implementation:  Training
The implementation phase of an ERP can take weeks, even
Implementation  The duration of ERP implementation
months, depending on the number of modules to be
integrated, the complexity of the business processes and the  The complexity of implementation
implementation methodology.  Successful references

3. APPLICATION
In this part, the AHP method will be applied to determine the
Phase 2: Define the ERP shortlist.
best ERP according to the criteria that will be determined in a ERP selection project, the selection team should identify
later. key business challenges that the enterprise aims to solve with
Before applying the AHP method, a set of the ERP selection the ERP system, gather the requirement of each department,
phases must be performed by the project team, the list of and identify the ERP system characteristics.
proposed selection phases is presented as follows: Another task of the team is the collection of the possible
information about the ERP systems and the vendors, and
 Define the evaluative criteria used to select the create an initial shortlist of vendors.
suitable ERP supplier. To obtain the ERP shortlist, the selection team should identify
 Define the ERP shortlist. ERP solutions specific to the company industry, and research
 Establish the AHP Model. the ERP system employed by similar companies.
 Application of the AHP Method an ERP selection team must have among its members an
expert consultant with extensive experience in ERP projects
Phase 1: Define the evaluative criteria used to select the as well as a clear vision of all the ERP solutions of the market,
suitable ERP supplier: its characteristics and its technical and functional
The ERP selection team is responsible for determining the specifications. this consultant, in collaboration with the
selection criteria according to the constraints of the enterprise different members of the team, can propose an initial list of
and the advice of the external consultants. ERP systems.
The criteria should be discussed with the company managers. In this example the shortlist is composed of two of ERP
According to these discussions the criteria are determined. In systems that called ERP A and ERP B.
this example the following criteria will be used:
 Adaptability criteria (C1). Phase 3: Establish the AHP Model:
 Financial criteria (C2). Hierarchical model of problem is composed according the
 Simplicity criteria (C3). criteria below (step 1). Level one represents the goal, which is
 Provider services (C4). to select the suitable ERP provider. The second level
 Implementation approach (C5). represents selection criteria, followed by the alternatives in
each of the above criteria has been decomposed into several the third level. (fig1)
sub-criteria to better understand the meaning of each criterion.
the sub-criteria have been given only for information, they
will not be taken into consideration in the various calculations
carried out by applying the AHP method. otherwise the AHP
method proposes directives which allow the weights of each
sub-criterion to be calculated using the local pairwise
comparisons and the weights of the criteria already calculated.

17
 International Journal of Computer Applications (0975 – 8887)
Volume 176 – No.7, October 2017

Fig 1: AHP Model

Phase 4: Application of the AHP Method C3 1/7 1/7 1 1/3 1/3

Each criterion will be evaluated compared to the others.
Usually, this step is obtained based on the views of the
selection team (following discussions at project team C4 1/5 1/5 3 1 1/3
meetings). For example: Adaptability is more important than
simplicity with 5 points(fig.2). C5 1/3 1/3 3 3 1

In order to obtain the weight of each criterion, the following

instructions are used:
First, the sum of each column is calculated, then the
normalization of the matrix by dividing the content of each
cell by the Sum of its column, and finally calculation of the
average of the rows (Table 2, Table 3 and Table 4) :

Table 3. The sum of columns

Fig 2: The pairwise comparison
Criteria C1 C2 C3 C4 C5
Based on the evaluations of previous step, we build the pairwise
comparison matrix according to the following rule: C1 1 3 7 5 3

C2 1/3 1 7 5 3
1
aij  0, a ji  , aii  1
aij C3 1/7 1/7 1 1/3 1/3

C4 1/5 1/5 3 1 1/3

The value of the (i,j) position of the pairwise comparison matrix
is determined using Saaty’s scale(1,3,5,7,9), the inverse value of
C5 1/3 1/3 3 3 1
the assigned number is assigned to the ( j,i ) position (fig.3).
Sum of
Table 2. The pairwise comparison matrix 2,01 4,68 21,00 14,33 7,67
column

Criteria C1 C2 C3 C4 C5
Table 4. Normalized pairwise comparison matrix
C1 1 3 7 5 3
Criteria C1 C2 C3 C4 C5
C2 1/3 1 7 5 3
C1 0,50 0,64 0,33 0,35 0,39

18
 International Journal of Computer Applications (0975 – 8887)
Volume 176 – No.7, October 2017

The final test is:

C2 0,17 0,21 0,33 0,35 0,39
If CR<0,1: then the matrix is consistent and judgments can be
C3 0,07 0,03 0,05 0,02 0,04 considered coherent.

C4 0,10 0,04 0,14 0,07 0,04 If CR>0,1: then the matrix results are inconsistent and were
exempted for the further analysis.
C5 0,17 0,07 0,14 0,21 0,13
The next step is to check the results of the previous example:
Table 5. Calculation of criteria’s weight  First the Weight sums vector Ws will be determined:
Sum of
Weight
Criteria C1 C2 C3 C4 C5 line  2, 45   1 3 7 5 3   0, 44 
  1/ 3 1 7 5 3 
  
2,21 0,44  1,57    0, 29 
 0, 22    1 / 7 1 / 7 1 / 3 1 / 3
C1 0,50 0,64 0,33 0,35 0,39
1  0, 04 
  1/ 5 1/ 5 3 1 1 / 3
  
 
1,45 0,29
C2 0,17 0,21 0,33 0,35 0,39
 0, 40  0, 08 
 0, 76   1 / 3 1 / 3 3 3 1   0,14 
   
C3 0,07 0,03 0,05 0,02 0,04 0,22 0,04
 Find the Consistency vector:

C4 0,10 0,04 0,14 0,07 0,04 0,40 0,08  5,53   2, 45   0,44 1


     1 
0,72 0,14  5, 40   1,57   0,29 
C5 0,17 0,07 0,14 0,21 0,13
 5,15    0, 22    0,04
1 

     1 
Conclusion: Adaptability is the most important criterion with a  5, 07   0, 40   0,08 
 5, 26   0, 76   1 
0.44 weight, then comes financial criterion with 0.29 weight, the
     0,14 
third criterion is the implementation approach with a 0,14 weight,
then comes provider services criterion with a weight of 0,08 and
 Find max and determinate CI and CR:
the last one is simplicity criterion with a weight of 0.04.

5,53  5, 40  5,15  5, 07  5, 26
Consistency check:
max   5, 28
Before proceeding to the next step of the AHP method, it is 5
essential to make sure that we did not make any absurd
comparisons and that criteria weights are indeed consistent, that’s
5, 28  5
why this step is mostly important, to check the system consistency CI   0, 07
it is necessary to follow the steps below: 4

 Calculate Weight sums vector: {Ws} = {M}.{W} 0, 07

1 } CR   0, 06
 Find the Consistency vector: {Cv} = {Ws}. { 1,12
W
 Average Consistency vector, this quantity is called CR<0.1 The consistency ratio is acceptable; the system is
λmax.
consistent.
 Determinate Consistency Index :

CI 
 max  n  Alternatives weights:
 n  1 This is the final step of AHP. The goal is to get the weight of each
CI ERP solution in order to choose the best ERP system, for each
 Calculate Consistency ratio : CR  criterion, the matrix of ERP solution will be created, there are five
RI criteria, that’s why five matrixes are obtained.
The RI or Random Index is obtained from the table below, the following the same steps as before, the ERP weights are
determined as shown below:
first row represents the number of criteria (n):
Table 6. Alternative’s weight for Adaptability
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Alternatives ERP A ERP B Weight
0.00 0.00 0.58 0.901.121.241.321.411.451.491.511.481.561.571.59 0,83
ERP A 1 5
Source: Saaty (1977) 0,17
ERP B 1/5 1

19
 International Journal of Computer Applications (0975 – 8887)
Volume 176 – No.7, October 2017

Table 7. Alternative’s weight for Financial: the ERP, criteria, sub-criteria whenever they exist, and
alternative solutions (ERP).
Alternatives ERP A ERP B Weight
The use of the AHP method requires firstly, the determination
1 1/5 0,17 of the list of criteria that will be used to evaluate the ERP,
ERP A
then elaboration of a shortlist of alternatives ERP, and
5 1 0,83 subsequently the weight of each criterion related to the others.
ERP B
AHP method uses pairwise comparison in order to determine
Table 8. Alternative’s weight for Simplicity: the weights of the criteria and to evaluate the ERP systems
(alternatives), these pairwise comparisons are performed by
Alternatives ERP A ERP B Weight decision makers.
1 1/3 0,25 The pairwise comparison use a Saaty scale which show
ERP A
decision makers preferences and judgments among the
3 1 0,75 following options: equal importance, moderate importance,
ERP B
strong importance, very strong importance, and absolute
Table 9. Alternative’s weight for Provider services: importance.

Alternatives ERP A ERP B Weight Decision makers could provide a paired comparison with a
precision that depends on multiple elements such as their
1 3 0,75 knowledge in the field of ERP, their experience in similar
ERP A
projects, and the understanding level of the needs of the
1/3 1 0,25 company and its constraints.
ERP B
Using AHP, does not guarantee finding the optimal solution,
Table 10. Alternative’s weight for Implementation approch: decision makers must understand details, strengths, and
limitations of AHP method.
Alternatives ERP A ERP B Weight
Even though the nine-pointe numerical scale has the
1 7 0,88 advantages of simplicity and easiness for use, it does not take
ERP A
into account the uncertainty related to the preferences of
1/7 1 0,13 decision makers. Indeed, some ERP selection criteria contain
ERP B
ambiguity and multiplicity of meaning (adaptability,
The results (weights) are combined in this matrix: simplicity...). Furthermore, the human assessment on
qualitative attributes is generally subjective and thus
Table 11. Matrix of weights: imprecise. For this reason, in some cases, AHP seems
inadequate to determine accurately the selection criteria
weights.
Alternatives C1 C2 C3 C4 C5 Fuzzy concept can be incorporated with the pairwise
comparison as an extension of AHP, in order to represent this
0,83 0,17 0,25 0,75 0,88 kind of uncertainly related to decision makers judgments.
ERP A
0,17 0,83 0,75 0,25 0,13 Thus, the triangular fuzzy numbers can be used to improve the
ERP B
nine-point scaling scheme and represent subjective pairwise
comparisons of ERP selection criteria, hence the use of Fuzzy
To obtain the last result the transpose of the matrix above is
AHP method.
multiplied by the weight vector determined above (Table 4):

 0, 44  5. REFERENCES
  [1] Moutaz Haddara and Ondrej Zach (2011). ERP Systems
0, 29 
 0, 61   0,83 0,17 0, 25 0, 75 0,88  
in SMEs: A Literature Review.
    0, 04  [2] Malhotra and Temponi. Critical decisions for ERP
 0, 38   0,17 0,83 0, 75 0, 25 0,13    integration: Small business issues, International Journal
 0, 08  of Information Management, Volume 30, Issue 1,
 0,14 
  February 2010, Pages 28–37.
[3] Huseyin Selcuk Kilic , Selim Zaim and Dursun Delen.
We finally establish the weight of each ERP, 0.61 for ERP A and Selecting “The Best” ERP system for SMEs using a
0.38 for ERP B, based on the five criteria and the data we combination of ANP and PROMETHEE methods, Expert
gathered, and by applying the AHP method, we can conclude that Systems with Applications, Volume 42, Issue 5, 1 April
the ERP A is more suitable for the company compared to the ERP 2015, Pages 2343–2352.
B. [4] Kilic, Zaim, and Delen. Development of a hybrid
methodology for ERP system selection: The case of
4. CONCLUSION & DISCUSSIONS Turkish Airlines, Decision Support Systems Journal,
MCDMs are often used in ERP selection process, AHP is Volume 66, October 2014, Pages 82–92.
considered among the most popular methods for this type of
[5] Gürbüz, Alptekin and Isıklar Alptekin. A hybrid MCDM
treatments. the application of the AHP method starts with a
methodology for ERP selection problem with interacting
problem modeling through a simple and clear hierarchy which
allows its decomposition into three or four levels: selection of

20
 International Journal of Computer Applications (0975 – 8887)
Volume 176 – No.7, October 2017

criteria, Decision Support Systems Journal, Volume 54, [14] Cebeci. Fuzzy AHP-based decision support system for
Issue 1, December 2012, Pages 206–214. selecting ERP systems in textile industry by using
balanced score card, Expert Systems with Applications,
[6] Lin, Chen, Ting. An ERP model for supplier selection in Volume 36, Issue 5, July 2009, Pages 8900-8909
electronics industry, Expert Systems with Applications
Journal, Volume 38, Issue 3, March 2011, Pages 1760– [15] C.Kahramane. Selection among ERP outsourcing
1765. alternatives using a fuzzy multi-criteria decision-making
methodology, International Journal of Production
[7] Perçin. Using the ANP approach in selecting and Research, Volume 48, 2010 - Issue 2.
benchmarking ERP systems, Benchmarking: An
International Journal, Vol. 15 Issue: 5, pp.630-649 [16] Karsak, Özogul. An integrated decision-making approach
(2008). for ERP system selection, Expert Systems with
Applications, Volume 36, Issue 1, January 2009, Pages
[8] Sandarbh Shukla, Mishra, Rajeev Jain, Yadav. An 660–667.
integrated decision-making approach for ERP system
selection using SWARA and PROMETHEE method, [17] Burak Efe. An integrated fuzzy multi criteria group
International Journal of Intelligent Enterprise, volume 3, decision making approach for ERP system selection,
Issue 2 (2016). Applied Soft Computing, Volume 38, January 2016,
Pages 106–117.
[9] Ayag˘ and Özdemir. An intelligent approach to ERP
software selection through fuzzy ANP, International [18] Maria Manuela Cruz-Cunha. ERP Selection using an
Journal of Production Research, Volume 45, Pages 2169- AHP-based Decision Support System, IRMJ journal,
2194, Issue 10, (2007). volume 29 , Issue 4 (2016).
[10] Wei, Chien and Wang. An AHP-based approach to ERP [19] Moutaz Haddara. ERP Selection: The Smart Way,
system selection, International Journal of Production Procedia Technology, Volume 16, 2014, Pages 394-403.
Economics, Volume 96, Issue 1, 18 April 2005, Pages
47–62, (2005). [20] Igor Rivera and María del Rosario Pérez Salazar (2013).
GUÍA DE SELECCIÓN DE ERP EN LAS PEQUEÑAS
[11] Özdağoğlu and al. Comparison Of AHP And Fuzzy AHP Y MEDIANAS EMPRESAS MEXICANAS
For The Multi-Criteria Decision Making Processes With
Linguistic Evaluations, 2007. [21] Wen-Hsien Tsai, Pei-Ling Lee, Yu-Shan Shen and Hsiu-
Ling Lin (2013). A comprehensive study of the
[12] Gunasekara, Ngai, and McGaughey. Information relationship between entrerprise resource planning
technology and systems justification: A review for selection criteria and enterprise resource planning
research and applications, European Journal of success.
Operational Research, Volume 173, Issue 3, 16
September 2006, Pages 957–983. [22] Huseyin Selcuk Kilic, Selim Zaim and Dursun Delen..
Selecting “The Best “ERP system for SMEs using a
[13] Ünal and Güner. Selection of ERP suppliers using AHP combination of ANP and PRPMETHEE methods.
tools in the clothing industry, International Journal of
Clothing Science and Technology, Vol. 21 Issue: 4,
(2009).

IJCATM : www.ijcaonline.org
21