You are on page 1of 10

Analysis of the decision process for selecting ERP

software: the case of Keller Manufacturing

Jacques Verville
Department of Accounting and Information Systems, College of Business
Administration, Texas A&M International University, Laredo, Texas, USA
Alannah Halingten
Department of Accounting and Information Systems, College of Business
Administration, Texas A&M International University, Laredo, Texas, USA


Software use,
Enterprise resource planning,
Systems integration


Keller Manufacturing, a mid-sized

furniture manufacturer, completed
the purchase of an enterprise
resource planning (ERP) software
solution in August of 1996 at a
cost of US$1 million. From 12
individuals who participated on
Keller's acquisition team, the four
principals were interviewed for
this case. The structure of the
acquisition process that emerged
from the data revealed six
distinctive iterative, recursive and
inter-related processes that,
together, form a complex web
of activity and tasks for the
acquisition of ERP software.
These activities and tasks are
described and analyzed as a
function of the six processes. The
ERP acquisition process
developed by Keller for this
purchase was atypical of their
normal purchasing practices and
proved to be a significant learning
experience for the entire
organization. This case provides a
useful illustration of ``good
practice'' and sets forth the
framework for the ERP acquisition

Received May 2001

Revised December 2001
Accepted January 2002

Integrated Manufacturing
14/5 [2003] 423-432
# MCB UP Limited
[ISSN 0957-6061]
[DOI 10.1108/09576060310477825]


the first order, to find out what indeed the

process is that organizations go through to
buy ERP software. As such, the research
questions for the study were: ``How do
organizations acquire packaged software?
What are the processes that are involved for
packaged software acquisitions?''
The focus of this paper, then, is on the
acquisition process that a medium-size
midwest furniture manufacturing company,
Keller Manufacturing, went through in 1996
for a manufacturing execution system (MES).
The paper will begin with a literature review
of the management information systems
(MIS) field and be followed by the research
methodology that was used for the study. The
latter section will include a review of the
methods used for data collection, data
analysis and ensuring the validity of the
study, as well as present the limits of the
study. Immediately following this section
will be Keller's corporate profile and a
description of the circumstances that led to
the decision to buy a packaged ERP software
solution. Subsequent to that, an overview of
the activities that Keller went through to
purchase ERP software will be presented. An
analysis of Keller's ERP acquisition process
will follow, along with the lessons that Keller
learned from this buying experience.

As a suite of integrated software application

modules that can link back-office operations
to front-office operations as well as internal
and external supply chains, enterprise
resource planning (ERP) software conjoins
functional areas and business processes in an
integrated environment that provides a
broad scope of applicability for
organizations. While considered a viable
althernative to in-house development
(Verville and Halingten, 2001; Verville, 2000;
Eckhouse, 1999; McNurlin and Sprague,
1998), the purchase of ERP software is not
without its challenges. At a cost of several
thousands, hundreds of thousands or even
millions of dollars, the purchase of ERP
software is a high-expenditure activity that
consumes a significant portion of their
capital budgets. It is also an activity that is
fraught with a high level of risk and
uncertainty. Why? Because, first of all, if a
wrong purchase is made, it can adversely
affect the organization as a whole, in several
different areas and on several different
levels, even to the point of jeopardizing the
very existence of the organization. This
highlights the obvious need for making the
right choice of software. It also brings to light
the need for finding the best means for
acquiring this type of software so that the
right choice can be made. Second, because of
the implementation and the risk of it going
awry. ERP implementations are said to be the
single business initiative most likely to go
wrong (Verville and Halingten, 2001;
Verville, 2000; Hill, 1999).
In light of these concerns, a research
project was undertaken to determine the best
way to acquire ERP software. However, with
little or no prior research found on the topic
of ERP acquisitions, it became necessary, in

A review of the literature in the field of MIS

shows that research conducted in the area of
ERPs has concentrated on implementation
and post-implementation issues (Esteves and
Pastor, 2001; Verville, 2000).
The type of problems and issues that arise
from the implementation of ERP systems
range from specific issues and problems that
can come up during the installation of an
ERP, to behavioral, procedural, political, and

The Emerald Research Register for this journal is available at

The current issue and full text archive of this journal is available at

Literature review

[ 423 ]

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

[ 424 ]

organizational changes, etc. that manifest

subsequent to the installation. One research
topic, that of critical issues that affect an ERP
implementation, is the focus of the study that
was conducted by Bingi et al. (1999). In a
similar study, Brown and Vessey (1999) focus
on implementation practices to uncover the
implementation variables that appear to be
critical to the success of ERP
implementations. Another research area is
that of organizational change. In this area,
Boudreau and Robey (1999) present a
framework to guide research on ERP-related
organizational transition (i.e. organizational
change as a process). Another study by Koh
et al. (2000) uses a framework, based on a
process theory approach, to understand and
explain the ERP implementation experiences
of organizations. Another subject of research
within the area of organizational change is
the roles of individuals within organizations.
In this vein, Caglio and Newman (1999)
propose a study aimed at understanding the
change in the roles of management
accountants within the wider socio-economic
context of the organization in which new
ERPs are introduced. Other interesting
topics for research include user buy-in,
commitment (management, team,
organization, etc.), leadership, organizational
culture, stakeholders, organizational learning
and communications, to name but a few
(Bingi et al., 1999; Boudreau and Robey, 1999;
Glover et al., 1999; Hill, 1999; Miranda, 1999;
Riper and Durham, 1999; Sieber et al., 1999;
Appleton, 1997; Best, 1997).
While much attention is directed to
implementation, post-implementation and
other organizational issues, the issue of the
acquisition process for ERP software is for
the most part being ignored. This issue is
important, however, because as the stage
preceeding the implementation process, it
presents the opportunity for both researchers
and practitioners to examine all of the
dimensions and implications (benefits, risks,
challenges, costs, etc.) of buying and
implementing ERP software prior to the
commitment of formidable amounts of
money, time and resources. For researchers,
the challenge would then be to ascertain the
correlation between the acquisition process
and the implementation process, the results
of which could be beneficial to practitioners.
Another objective of the research presented
herein was to gain an understanding of
current organizational ERP acquisition
practices. The light shed on this process may
help practitioners and contribute to the
change and improvement of present ERP
acquisition practices.

Research methodology
Owing to the nature of the study, the
research strategy was to do a multiple-case
design involving four organizations (one of
which is the subject of this paper) from
different business sectors (manufacturing,
telecommunications, transportation, energy)
that had recently completed the acquisition
of new ERP software. Since the area of ERP
acquisition is a relatively new area of
research, the case study approach provides
the best means to start exploring this issue.
This approach was thought to be particularly
well suited for this study because it was
expected to unveil a multitude of factors and
dimensions that make the acquisition of ERP
software such a complex process. The
rationale for the multiple-case design was
that as a research strategy, we could direct
our focus to understanding the dynamics and
complexities present within each case (Yin,
1989; Miles and Huberman, 1994), these being
the processes and critical issues of software
acquisition within organizations.
Site selection for the study was made
according to the following five criteria:
1 the acquisition had a significant impact
on the organization;
2 the acquisition was significant, totaling
several hundred thousand dollars or
3 the type of packaged solution that was
acquired was of a complex nature such as
4 the acquisition was a new purchase; and
5 the acquisition of the software was
recently completed.

Data collection

Data collection was done in three parts. The

first part consisted of semi-structured
interviews. Interviews were conducted with
four individuals, each lasting approximately
1hr and 15min. The informants, all of whom
were directly involved in the acquisition
process, included the VP of information
systems, the VP of personnel, the corporate
materials manager, and a plant manager.
Open-ended questions were used
throughout the interviews. They allowed for
flexibility and provided the:

. . . possibilities of depth; they [also] enable[d]

the interviewer to clear up
misunderstanding[s] (through probing), to
ascertain a respondent's lack of knowledge, to
detect ambiguity, to encourage cooperation
and achieve rapport, and to make better
estimates of the respondent's true intentions,
beliefs, and attitudes (Kerlinger, 1986, p. 442-3).

As it so happened, the informants sometimes

gave unexpected answers that indicated the

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

existence of relations (activities, tasks, and

influences) that were not originally
anticipated and this added to the richness of
the cases.
For this study, the opening question for the
interview with each informant was as
In your own words, how did you go about
buying packaged ERP software? Describe how
the acquisition went, from the beginning to
the end.

Following the informant's description,

follow-up probing and topical questions such
as ``Was there a planning scenario? If so, in
your own words describe the planning aspect
of the process. Describe in your own words
the various aspects of the packaged software
acquisition planning scenario'' were used to
clarify an issue or to delve for more
information. These follow-up questions also
allowed for the development of ideas without
constraining the exploratory nature of the
study. The same interviewing protocol was
observed with all of the informants.
The second part of the data collection
consisted of gathering archival information
from various sources within the organization
and included documentation from
acquisition project, plans, designs, best
practices, policies, request for information
(RFI), letters and memos, reports, etc. These
documents allowed for a closer examination
of what happened during the decision
For the third part of the data collection
member checks were conducted. Participants
were asked to review their transcripts from
the interviews in order to verify content, and
when necessary, amend or add to them.
Follow-ups were also conducted by telephone
and/or e-mail to clarify ambiguities or
discrepancies, or to confirm information.
Feedback from others (peers independent of
the organizations and individuals who
participated in this study) was also obtained.


All interviews were audio-taped for

subsequent transcription and for verification
of accurate interpretation. Member checks
were performed (as described in the previous
section). Feedback was also obtained from
individuals independent of the study.
The data from this study was validated
using a triangulation method. According to
Robson (1993, p. 383):
. . . the by-products of triangulation are as
useful as its primary purpose in validating
information. It improves the quality of data
and in consequences the accuracy of the

In addition, validity, according to Maxwell

(1996, p. 87), refers to the:

. . . correctness or credibility of a description,

conclusion, explanation, or other sort of

For example, the triangulation of data

sources within one case was repeated in each
of the other three cases and then for all of the
cases together. The results show that while
each of the cases is different with regard to
the type of software solution that was being
acquired, the same process was developed
and similar tasks were performed.

Limits of the study

The limitations of this study can be linked to

the choices that were made regarding the
research and specifically relate to the
newness of the research topic, that being the
acquisition of ERP software, the minimal
amount of research that has been conducted
to date in this area, and the methodology that
was used for the study.
Given the lack of literature on this specific
subject, the case study approach was selected
as the best means to gain the maximum
knowledge and understanding about
packaged software acquisition activities,
issues, dynamics and complexities. However,
each method has its strengths and
drawbacks, and the case study approach is no
exception. It is more limited than surveys in
terms of generalizability. While surveys
enable precise extrapolation of results to a
defined population (Maxwell, 1996), case
studies are more limited in their focus. As
such, a single or a few cases are poor
representations of a population of cases and
may be poor grounds for generalization. This
having been said, a single case as a negative
example can establish limits to grand
generalization (Maxwell, 1996; Yin, 1989).
Hence, case studies are of value in refining
theory and suggesting complexities for
further investigation, as well as helping to
establish the limits of generalization.
In the next section, we will be presenting
the organization, background information
about the case, the case itself and an analysis
of each of the processes that comprise the
ERP acquisition process beginning with
planning, followed by information search,
selection, evaluation, negotiation and choice.

Keller Manufacturing Company:

organizational profile
Keller Manufacturing Company was
established in 1895 as a manufacturer of farm
wagons and did so until 1943 when it began
manufacturing household furniture. Today,

[ 425 ]

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

this organization has over 700 employees in

three manufacturing plants in the USA (two
of them located in Indiana (Corydon and New
Salisbury) and one in Culpepper, Virginia)
and manufactures over 2,000 different oak
and maple legs, seats, and other components
(with over 100 separate procedures) that are
required in the assemblage of its products. In
1995, the company earned profits of $3.1
million on sales of $46 million. This
represented a 76 per cent increase in profits
with only a 30 per cent increase in sales from
the previous year. In three short years,
Keller Manufacturing grew from $35 million
in sales to $54 million in 1996, representing a
54 per cent increase in sales.

Background leading to Keller's decision to

buy an ERP

During the last few years, Keller changed

from a production-oriented company to a
very effective market-driven business. It also
expanded its product-line to include bedroom
furniture, a change that proved to be a very
successful marketing strategy. Consequently,
the re-orientation of Keller's marketing
strategy with the resulting increase in sales
and new product introductions resulted in
some production challenges to their
manufacturing operation. Manufacturing
was having difficulty supporting production
demands brought on by the substantial
increase in sales. Mainly, they attributed the
problem to the lack of timely and accurate
information that was necessary to effectively
and efficiently plan for and control
At the time, Keller's information systems
consisted of a combination of manual
procedures and automated systems with
computers only being used by the engineers
and in manufacturing. Keller had one main
computer, an AS400, that was used for
batch-oriented processing and several
stand-alone PCs. These stand-alone systems
were originally installed in an effort to
provide information that was desperately
needed to support manufacturing operations.
However, with the recent and sizable
increase in sales and sales mix (expanded
product-line), this type of information system
(manual and computerized) could no longer
effectively support Keller's manufacturing
Unfortunately, the weakness of this
information system was not limited to
Keller's manufacturing operations.
Personnel in other areas of the organization
were having difficulty performing their
duties effectively because of the
inefficiencies of the manual systems

[ 426 ]

As sales increased, the organization's

ability to control costs and provide
customers with products on a timely basis
became more and more difficult. Competition
continued to shorten lead times. Keller was
faced with an escalating problem that needed
to be urgently addressed the organization
could no longer continue status quo and still
remain competitive and profitable. They
were forced to seriously consider more
``modern'' systems that could help them
better utilize their personnel, equipment, and
facilities, and provide better service to their
Hence, in 1995, Keller Manufacturing
requested the assistance of an outside
state-run agency to conduct an assessment of
its organization. The objective of this study
was to identify areas in which modern
manufacturing techniques and new
technology could benefit the organization in
terms of performance and competitiveness.
The study showed that Keller was deficient
in several areas and, in particular,
manufacturing. The report outlined that
Keller's manufacturing operations were
controlled by a combination of manual and
automated procedures and stand-alone PCs.
However, with the growth that they were
experiencing, the existing systems and
procedures could no longer support
operations and provide management with the
information that it needed. If no action was to
be taken, the organization's continued
growth and ability to support its customer
base would be compromised.
The recommendation was therefore made
that Keller invest in a fully integrated
manufacturing system, more specifically, a
multi-user, multi-tasking, fully integrated,
real-time manufacturing resource planning
system. The study showed that
implementation of a resource planning
system could improve scheduling, efficiency
in the assembly area (by keeping the lines
running), tracking, security, controls over
shipping areas, manufacturing processes and
individual performance; it would reduce
setup times and thereby improve efficiency,
reduce unreported parts losses, and reduce
database and key-entry errors. New
employees could also be trained more
efficiently and effectively with the new
system. It was estimated that the new system
would save Keller approximately $1.2 million
In October 1995, with the realization that
their existing system could no longer sustain
the current rate of growth and in order to
take full advantage of future market trends,
Keller's senior management decided that the
company would acquire a manufacturing

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

execution system (MES), or in effect, an

extensive information system for the shop

Description of purchasing activities for

ERP software

Once the decision was made to proceed with

the purchase of a new manufacturing system,
meetings were held and a team of 12
individuals was selected to be part of the
buying process. The various areas
represented on this team were finance,
personnel, marketing, engineering,
production (manufacturing and operations)
and information systems. At the outset, it
was decided that the team's task would be to
find the very best integrated solution not
only to meet Keller's immediate needs in
manufacturing (on the shop floor), but also
its long-term organizational needs. It was
also decided that the selection of the system
would be by the users. With this in mind,
meetings with the line managers' at all three
plants were arranged during which the team
compiled a ``wish list'' of requirements for the
new system. An employee advisory
committee (EAC) was also formed at about
this time that brought representatives from
various user groups together. Since Keller
was going to be introducing numerous
changes to the organization, they realized
that user buy-in from the various sectors
would be very important to the successful
outcome of the project. In addition to the
EAC, Keller wanted to make certain that all
employees were kept abreast of how things
were progressing with the acquisition.
Hence, weekly notices were published and
distributed to all employees to keep them
informed throughout the entire process.
In the initial phase of the acquisition
process, certain members of the team began
looking for companies that had developed
manufacturing execution systems. Their
principal reference was Manufacturing
Systems Magazine in which was published a
list of the top 50 manufacturing software
developers/vendors. From this list, they
selected the vendors that had manufacturing
execution systems (MES) and/or ERP
(enterprise resource planning) software.
Then, they contacted the MESA
(Manufacturing Execution Systems
Association) Group which is an association
of 11 member organizations. From this, they
created their long-list of 62 vendors and
subsequently compiled information on each
of them.
At this point, a ``selection team'' of five
individuals was formed from the initial
12-member acquisition team. These five
individuals took on the task of selecting and

evaluating potential vendors and their

products. To assist them with this task, they
refined the line managers' ``wish list'' and
then expanded it to what became their list of
critical requirements. This list was then used
as one of their tools of evaluation of the
various vendors' products. The ``selection
team'' proceeded to review the information
that had been gathered on the 62 vendors and
eliminated 32 of them from further
consideration. A subsequent, more in-depth
evaluation was then done on the remaining
20 vendors that eliminated another 15 of them
from the team's list. The remaining 15
vendors were each sent a letter along with
the list of critical requirements that Keller
wanted the system to be able to do. After the
team had received and evaluated the
vendors' replies, each of the 15 vendors was
invited to Keller's head office to meet with
the acquisition team and present their
software. Of the 15 vendors, seven did not
respond, three submitted reports on their
products and five went to meet with Keller's
team for a preliminary meeting and sales
presentation. The five vendors that met with
Keller's team were evaluated based on their
demonstrations and the supplemental
information that the selection team had
gathered on them. At this point, the team also
ran Dunn & Bradstreet (D&B's) financial
reports on each of the eight vendors (the
three that submitted reports and the five that
did presentations). The team used these
reports to rank each vendor based on their
number of employees, financial stability,
product, and other criteria. Further to this,
the team was able to conduct six site visits to
various vendor locations and customer
installations (not all of the vendors were
willing to supply referrals). Additional
references were also requested which they
subsequently contacted. Some of the
information that they received from these
referrals provided the selection team with
more insight into what they should be
looking for, and they used this information to
further refine their list of critical
After reviewing the D&B reports, their
impressions from the demonstrations, the
input that they had received from the site
visits and the referrals, along with their
critical requirements list, they reduced their
list of eight vendors to a short-list of three
Keller then developed a comprehensive
listing of specific performance criteria from
their list of critical requirements. The
selection team then used these requirements
to evaluate vendor and software capability in
detail. Further, Keller visited five companies

[ 427 ]

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

which were using the systems from their

short-listed vendors. Base on these
evaluations (using their comprehensive
listing of performance criteria and what they
found during the five site visits), they found
EMS (Effective Management Systems) to be
the most capable to fulfill their information
systems needs. Three of the selection team
members then spent two days with EMS
during which they learned more about their
company, their strategic direction and their
philosophies. All-in-all, the selection team
determined that EMS would be a good
long-term partner for them.
Keller then invited EMS back for intensive
two-day demonstrations for the selection
team, mill managers, and the EAC, a total of
approximately 50 to 60 individuals. These
individuals were then polled regarding their
impressions of the demonstrations and the
responses were overwhelmingly positive.
The decision was then made to select EMS's
packaged software solution.

Analysis of Keller's ERP

acquisition process
Analysis of Keller's activities to purchase
ERP software revealed several distinctive
elements, that have been grouped as
``processes'', that Keller completed for its
acquisition of a manufacturing execution
system. Indeed, six processes were found to
constitute Keller's ERP acquisition process
(ERPAP) and they are planning, information
search, selection, evaluation, choice and
negotiation. The analysis of each process is
presented below.
``Process'', as used herein, does not refer to
the flow of activities that one commonly
associates with process. Rather, process, in
the context of this study, denotes a grouping
of events/activities/tasks that occurred
during the acquisition process without
regard to the sequence in which the activities
took place. The reason for this is that the
analysis of the flow revealed iterative and
recursive activity within and between the
processes, that the processes sometimes
occurred concurrently, and that elements of
some processes made those processes appear
embedded in whole or in part in other
processes. While a high-level view of Keller's
activities could lead one to believe that the
ERP acquisition process is linear or
sequential in nature and could be easily
flow-charted, analysis of the data revealed an
oftentimes complicated web of activity that
was neither easy to classify nor describe. The
data did reveal, however, that at certain
points in each process when all requirements

[ 428 ]

for said process were ``believed'' to have been

completed, that a progression was made by
the team from one process to another. The
order of that progression is the basis for the
sequence in which the processes are
presented below. It needs to be noted by the
reader, however, that for the most part, the
processes and the activities of which they are
comprised did not occur sequentially.

Planning process

For Keller, planning marked the beginning of

the acquisition process. It was noted
within the data that ``planning'' encompassed
all of the activities that Keller deemed
necessary to pursue this endeavor. Keller's
planning process included meetings to
determine schedules, priorities, participants;
activities and tasks that would need to be
completed; types and sources of information
to be sought; and so forth. Even so, Keller's
planning process did not appear to have been
Although the term ``planning'' was not
specifically used in the responses to the
general question, it was used in the responses
to the question concerning the planning
process. In their responses to the general
question, the interviewees spoke of tasks and
activities that needed to be accomplished
such as the development of a list of
requirements and meetings with internal
customers who would be affected by the new
systems. According to Keller's VP of
information systems, ``. . . a planning scenario
as far as timing and dating, defining
milestones'' was developed and laid out in a
Gantt Chart using Microsoft Project
As just discussed, data from the interviews
and the supporting documentation showed
that a planning phase was completed by
Keller's acquisition team. What could also be
deduced from the data was that although the
level of planning may not have been as
extensive as it could have been (based on the
type and magnitude of this acquisition a
first for Keller), planning was continuous
throughout Keller's acquisition process.
Planning was done iteratively as the
acquisition process progressed, with the plan
adjusting as warranted by ``new'' information
as it was received.

Information search

What was most significant in the information

search process was the ``sources of
information''. We identified to two sources of
information: internal and external. In their
responses, the participants informed us that
information was obtained from both sources.

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

As to the internal information sources,

Keller availed themselves of information
from various sources within the organization
that included individual users and team
members. These internal sources provided
information primarily on the organization's
requirements (existing) at all of the levels
and in all of the areas that the technology
would impact.
External sources were sought to provide
information about software solutions that
might best meet their needs. Keller
conducted a marketplace search, gathering
information from competitors, trade
publications, journals, a seminar (addressing
re-engineering in the furniture industry with
emphasis on information and technology),
and professional associations (four
professional associations were contacted,
AFMA (American Furniture Manufacturers
Association), SME (Society of Manufacturing
Engineers), MESA (Manufacturing
Execution Systems Association) and APICS
(American Production and Inventory Control
Society)). According to the VP of information
systems, Keller approached MESA to obtain
information on the various members that
provided manufacturing execution systems.
Information gathered from these sources was
used by Keller in the construct of their
long-list of vendors.
In addition to the external sources listed
above, Keller also gathered information from
the visits that they conducted to some of the
vendors' sites and from vendor references.

Selection process

Concurrent to the planning process, several

iterations of screenings (which we
considered as part of the selection process)
were done during the information search
process prior to arriving at a shorter long-list
of vendors. Selection and evaluation criteria
pertaining to both the vendors and their
technologies were used to screen for vendors
who could supply the type of software
solution that Keller was seeking.
For the most part, we see the selection
process as having begun at the point when
Keller received the RFI responses back from
the vendors. According to the corporate
materials manager, Keller's selection process
was conducted in two phases. The first phase
was conducted upon receipt of the RFIs from
the vendors. With the RFI responses in hand,
Keller proceeded with the ``paper'' evaluation
of the vendors' packages, that is, they
evaluated the responses as they were
presented, at face value, on paper. More
precisely, this entailed an initial cursory
evaluation that was done by the VP of
information systems that quickly eliminated

more than half of the 62 vendors on the

long-list. At this point, the second phase of
the selection process began. A second contact
was made with the remaining 30 vendors by
way of a letter (a second RFI) that detailed the
functionalities and capabilities Keller was
seeking from the vendors and their software.
Following an in-depth evaluation of the
responses that were received, 15 vendors/
software were eliminated from further
consideration. According to the VP of
information systems, Keller invited the
remaining 15 vendors to their facilities to
conduct in-house demonstrations (sales
presentations) of their proposed solutions. Of
the 15 vendors that were invited, only eight
vendors responded to Keller's invitation five
of the vendors went to Keller's facilities and
did in-house sales/product presentations,
while three submitted reports on their
products. Along with the evaluations that
Keller did of the in-house presentations and
of the reports that they received, all eight
vendors were evaluated using Dunn &
Bradstreet reports.
Among the factors that Keller considered
important in its selection of vendors were
financial stability and size. Another critical
factor was the ability of the vendor to meet
Keller's requirements. Keller wanted a single
vendor solution. Strategically, this was an
important factor for Keller. Although this
project was focused on manufacturing, Keller
was also thinking of their future
requirements in other areas of the
organization such as finance.
The data shows that information was also
gathered on some of the vendors and their
software solutions (functional aspects) by
means of reference site visits. Three
reference sites were visited by some of the
members of the acquisition team and all were
found to be valuable sources of information.
According to the VP of personnel, the
objective of the selection process was to
create a short-list of vendors who would later
be invited to do scripted in-house

Evaluation process

Keller's evaluation activities, as noted from

the data, fell into two distinctive categories,
vendor and functionality, and evaluation
criteria for both were developed in the
planning phase of the acquisition process.
The acquisition team assigned weights to
each of the criteria and ranked them. These
criteria were used to develop questionnaires,
evaluation matrices, and demonstration
scenarios for the different stages of the
evaluation process.

[ 429 ]

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

One of the matrices that was developed

(called ``critical requirements score card
(CRSC)'') was used during the selection
process for evaluating the RFI responses.
This score card contained 14 high-level
criteria for evaluating the functional aspects
of the software packages. The 14 criteria were
as follows:
1 receive information;
2 verify cut quantity;
3 schedule operations;
4 monitor schedule execution;
5 verify shop data input;
6 verify parts loss reporting;
7 detect labor variances;
8 provide real time status;
9 provide capacity planning;
10 calculate incentive pay;
11 verify inventory data entry;
12 provide needed flexibility;
13 provide operation history; and
14 provide security.
These criteria were used by the acquisition
team to develop mid-level and detailed
low-level criteria for measuring the vendors'
software. Keller used the detailed low-level
criteria to create a script for the in-house
product demonstration that it had its
primary vendor conduct as part of the last
stage of the evaluation process.
For Keller's vendor evaluation, each of the
vendors was evaluated in terms of their
financial stability, size, etc. based on reports
from Dunn & Bradstreet as well as other
information. In addition to the quantifiable
factors (sales volume, the size of the
company, etc.), according to the VP of
personnel, consideration was also given to
qualitative factors such as the quality of the
response, the appropriateness of the response
to Keller's particular needs, as well as the
impressions made during ``face-to-face''
meetings with the vendors.
For Keller's functional evaluation, high
and mid-level ``paper'' evaluations of the
software's functionality were done on the
vendors' responses to the first and second
RFIs. The team also evaluated software
functionality during standard or ``canned''
software demonstrations, intensive in-house
scripted product demonstrations, and visits
and calls to vendors' reference sites.
According to the VP of information systems,
various members of the acquisition team
visited certain vendor reference sites that
had purchased the technology Keller was
interested in. They also called the various
references that the vendors had provided.
The information and impressions that they
had gathered from these sources, together
with the in-house meetings and vendors'

[ 430 ]

software presentations enabled the

acquisition team to narrow its short-list of
eight vendors to a manageable few, from
which it singled out a primary candidate.
Analysis of the evaluation process revealed
much concurrent as well as iterative and
recursive activity between it and other
ERPAP processes, namely the planning,
information search, and selection processes.
Keller's vendor evaluation process occurred
in stages that were concurrent with other
processes, depending on the level of
evaluation that was required. For example,
Keller carried out high-level evaluation
activities during the information search
process to identify prospective vendors and
software solutions. During the selection
process, lower mid-level evaluations
occurred for the ``paper'' evaluations of the
vendors' RFI responses and during the
reference site visits. Dunn & Bradstreet
reports on the vendors were also evaluated at
this stage of the ERPAP. Iterative and
recursive activity occurred as new
information that was gathered by the
acquisition team forced it to revisit certain
processes, make adjustments (as required) to
plans, and then return/continue on with
vendor evaluation activities.
The same was true for the functional
evaluation process. Since the Keller
organization had little prior exposure to the
technology that it was seeking to implement
in its organization, the members of the
acquisition team were always learning and
adjusting to the information that it was
receiving. With this ``learn as you go''
approach, they built their list of functional
software requirements over time, integrating
what they knew they wanted the software to
do, with what they saw the software could or
could not do based on the vendors'
presentations, visits to vendor reference
sites, and so on. As such, the functional
evaluation process required several
iterations through parts of the planning
process, the information search process, the
selection process, and the choice process,
with additional sporadic returns as needed.
These occurred even as the acquisition team
was in the last stage of the functional
evaluation process. At this point in the
process, the acquisition team invited its
primary vendor to perform in-house scripted
software demonstrations.

Choice process

The acquisition process culminates in the

choice process and consists of the ``final
choice'' or ``recommendation'. In Keller's
case, a ``final choice'' or ``recommendation'' of
the software solution was arrived at by the

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller
Integrated Manufacturing
14/5 [2003] 423-432

acquisition team. The responses to both the

first and the second RFIs, meetings with the
vendors, calls to vendor references, site
visits, and standard/``canned'' software
presentations all contributed information
that enabled the acquisition team to narrow
their lists of potential vendors to a primary
choice. Then, for the purpose of confirming
their choice, Keller invited the primary
vendor in-house to do a scripted
According to the plant manager, the
acquisition team chose this particular
vendor's software solution because it was the
one (in the price range that Keller was
looking at and for the size of company that
Keller is) best able to meet Keller's current
and future requirements, not only on the
manufacturing side, but also in other areas of
the company.
In addition to the CRSC that was used
during the evaluation process, a meeting was
convened of all of the steering committee and
acquisition team members to see if there was
a consensus on which technological solution
was the most appropriate for Keller. Input
from the shop floor users who had
participated in the evaluation process was
also factored into the final choice.
Although the final recommendation rested
with the VP of information systems, it was ``a
group process'' (VP of personnel) and he (the
VP of information systems) ``looked at all of
us [as] having to agree on it''.

Negotiation process

It became obvious to us, even while

interviewing Keller's participants, that two
types of negotiations transpired during the
course of Keller's acquisition process. As the
``primary negotiator'', the VP of information
systems conducted business negotiations with
the vendors throughout the whole process on
such things as cost and modifications.
He also conducted legal negotiations on the
pricing and fine contractual details leading to
the signing of the contract with the vendor.
The business negotiations were
characterized as being informal, while the
legal negotiations were characterized as
formal. According to the VP of personnel,
they entered formal negotiations with their
vendor of choice, EMS, after the final choice
(recommendation) was approved by the
Board of Directors.
As to how this process could be
characterized, it appeared to us that the
business negotiations were fluid throughout
most of the process. As per the VP of
information systems, Keller did ``not pass up
an opportunity to negotiate as [they went]

Lessons learned
For Keller, the ERP acquisition process was
an incredible ``learning experience''. As
described by the VP of information systems:

. . . all of us learned unbelievably from this

experience. There was continual learning. We
were running into phrases and statements
that we'd never heard of before, especially
since none of us were really trained MIS

Among the lessons that this experience

taught them was that:
They could do it! Even though they took
the long route in doing things, they
learned that they could do it, on their own,
without the help of outside consultants.
They could adapt and accommodate a lot
of new information about technologies
that they knew little or nothing about.
This information was not always easy for
them to understand nor was it always ``cut
`n' dry''.
They had to develop new procedures for
dealing with ERP software acquisitions.
Since they had never previously
undertaken a process for this type of
software nor did they have in-house
expertise in this area, the acquisition
team had to ``construct'' (``design'') the
process as they went along. While there
were some fundamental purchasing
processes (minimal) in place for the
acquisition team to follow, in large part
(say 99 per cent), procedures had to be
developed ad hoc to deal with this ``new''
buying situation. This really was a ``learn
as you go'' process for Keller.
They had to modify their existing
purchasing procedures. Since Keller's
standard purchasing procedures were
inadequate, they had to be substantially
modified to suit the complexity of this
acquisition. With this being a first-time
purchase of packaged software of this
complexity and magnitude, beyond the
standard practice followed for the issue of
a purchase order, Keller did not have any
buying procedures in place that the
acquisition team could follow for this type
of purchase.
The VP of information systems carefully
documented the new procedures as well as
this process so as not to lose what the
organization had learned. Since this was a
new experience, the VP of information
systems wanted to leave behind some
documentation explaining the rationale that
was used and the decisions that were made
for choosing this (the EMS software)
technology over another. As a result, the

[ 431 ]

Jacques Verville and

Alannah Halingten
Analysis of the decision
process for selecting ERP
software: the case of Keller

acquisition team contributed a documented

history of what transpired during this
acquisition process to the organizational

Integrated Manufacturing
14/5 [2003] 423-432


[ 432 ]

Appleton, E.L. (1997), ``How to survive ERP'',

Datamation, Vol. 43 No. 3, pp. 50-3.
Best, C. (1997), ``Integrated system built on human
foundation'', Computing Canada, Vol. 23
No. 25, p. 54.
Bingi, P., Sharma, M.K. and Godla, J.K. (1999),
``Critical issues affecting an ERP
implementation'', Information Systems
Management, Summer, pp. 7-14.
Boudreau, M.C. and Robey, D. (1999),
``Organizational transition to enterprise
resource planning systems: theoretical
choices for process research'', Proceeding
ICIS, Charlotte, NC, pp. 291-9.
Brown, C. and Vessey, I. (1999), ``ERP
implementation approaches: toward a
contingency framework'', Proceeding ICIS,
Charlotte, NC, pp. 411-16.
Caglio, A. and Newman, M. (1999), ``Implementing
enterprise resource planning systems:
implications for management accountants'',
Proceeding ICIS, Charlotte, NC, pp. 405-10.
Eckhouse, J. (1999), ``ERP vendors plot a
comeback'', Information Week, No. 718,
pp. 126-8.
Esteves, J. and Pastor, J. (2001), ``Enterprise
resource planning systems research: an
annotated bibliography'', Communication of
AIS, Vol. 7 No. 8, pp. 1-52.
Glover, S.M., Prawitt, D.F. and Romney, M.B.
(1999), ``Implementing ERP'', Internal Auditor,
Vol. 56 No. 1, February, pp. 40-7.
Hill, S. Jr (1999), ``It just takes work'',
Manufacturing Systems, Selecting and
Implementing Enterprise Solutions
Supplement, pp. A2-A10.

Kerlinger, F.N. (1986), Foundations of Behavioral

Research, 3rd ed., Holt, Rinehart and Winston,
San Diego, CA.
Koh, C., Soh, C. and Markus, L. (2000), ``A process
theory approach to analyzing ERP
implementation and impacts: the case of
Revel Asia'', Journal of Information
Technology Cases and Applications, Vol. 2
No. 1, pp. 4-23.
McNurlin, B.C. and Sprague, R.H. Jr (1998),
Information Systems Management in Practice,
4th e., Prentice-Hall, Englewood Cliffs, NJ.
Maxwell, J.A. (1996), Qualitative Research Design:
An Interactive Approach, Sage Publications,
Thousand Oaks, CA.
Miles, M.B. and Huberman, A.M. (1994),
Qualitative Data Analysis: An Expanded
Sourcebook, 2nd ed., Sage Publications,
Thousand Oaks, CA.
Miranda, R. (1999), ``The rise of ERP technology in
the public sector'', Government Finance
Review, Vol. 15 No. 4, pp. 9-17.
Riper, K. and Durham, M.J. (1999), ``Phased ERP
implementation: the City of Des Moines
experience'', Government Finance Review,
pp. 37-42.
Robson, C. (1993), Real World Research: A
Resource for Social Scientists and
Practitioner-Researchers, Blackwell, Oxford.
Sieber, T., Siau, K., Nah, F. and Sieber, M. (1999),
``Implementing SAP R/3 at the University of
Nebraska'', Proceeding ICIS, Charlotte, NC,
pp. 629-49.
Verville, J.C. (2000), ``An empirical study of
organizational buying behavior: a critical
investigation of the acquisition of `ERP
software''', dissertation, Universite Laval,
Verville, J. and Halingten, A. (2001), Acquiring
Enterprise Software: Beating the Vendors at
Their Own Game, Prentice-Hall, Englewood
Cliffs, NJ.
Yin, R.K. (1989), Case Study Research: Design and
Methods, Sage, London.