Computers in Industry 56 (2005) 457471

www.elsevier.com/locate/compind

Best practices in business process redesign:

validation of a redesign framework
Selma Limam Mansar a,1, Hajo A. Reijers b,*
b

a
College of Business Sciences, Zayed University, PO Box 19282, Dubai, UAE
Department of Technology Management, Information Systems Group, Eindhoven University of Technology,
PO Box 513, NL-5600 MB, Eindhoven, The Netherlands

Received 22 July 2004; accepted 13 January 2005

Available online 2 March 2005

Abstract
A fundamental challenge in any Business Process Redesign (BPR) project is to come up with a new process design that is in
one or more ways superior to the existing plan. Based on earlier research, a framework to help the designer in selecting the proper
best practice(s) for this purpose is presented and validated in this paper. It is described how the framework is used in generating
improved process designs for two Dutch organisations. Furthermore, the results from a survey are presented, which has been
carried out among BPR practitioners in the UK and the Netherlands to test the framework. The overall conclusion is that the
framework is indeed helpful in supporting process redesign and that its core elements are recognised and put in practice by the
BPR practitioner community. The framework, therefore, may be of direct interest to both academics and practitioners active in
the process improvement field.
# 2005 Elsevier B.V. All rights reserved.
Keywords: BPR; Case study; Survey; Framework; Best practices

1. Introduction
Although overhyped, sharply criticised, and rebranded repeatedly over the past 15 years, business
processes redesign (BPR) has remained on the agenda
of many organisations. The simple reason is that it is
* Corresponding author. Tel.: +31 40 247 3629/2290;
fax: +31 40 243 2612.
E-mail addresses: selma.limammansar@zu.ac.ae (S.L. Mansar),
h.a.reijers@tm.tue.nl (H.A. Reijers).
1
Tel.: +971 4 2082 443; fax: +971 4 2082 653.

one of the most powerful ways to boost business

performance and raise customer satisfaction. Currently, in industry and the academic world there is a
growing interest for a field that is referred to as
Business Process Management [1]. Because Business Process Management is all about transferring the
results of BPR into production [2] it is more than
probable that the interest for what is the essence of
BPR, inventing new processes to do business, will rise
again in the coming years.
This paper aims at defining a framework to help the
process designer in choosing the correct best practice

0166-3615/$ see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.compind.2005.01.001

458

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

when trying to deliver a process design that is in some

sense superior to the existing one. We will refer to this
task as the implementation of BPR, to distinguish it
from other important BPR aspects and phases (e.g.
project and change management). The framework also
lists and classifies a set of best practices in BPR. The
idea behind a framework is to help practitioners by
identifying the topics that should be considered and
how these topics are related [3]. It is not a model of a
business process, but rather an explicit set of ideas that
helps in thinking about the business process in the
context of redesigning.
To define such a framework, we first have
investigated existing frameworks in literature and
adapted them for the specific purpose of BPR projects.
Next, we have applied the framework and the set of
best practices in redesign studies for two service
organisations, as presented in this paper. The results
led to a preliminary validation of the framework and a
classification of most used best practices. Finally, we
have conducted a survey amongst Dutch and UK
practitioners and experts in the field of BPR that
helped reconsider and validate our findings.
The organisation of this paper is as follows. The
first section describes our initial framework, the set of
best practices, and some open questions for these
subjects. The second section illustrates their application to two organisations and the conclusions drawn.
The third section provides findings from the conducted
survey on BPR. Finally, the conclusion describes the
validated framework and best practices and discusses
further implications and research.

Sundarajan [7]. In our framework, six elements are

linked (refer to Fig. 1):
 the internal or external customers of the business
process;
 the products (or services) generated by the business
process;
 the business process with two views:
a. the operation view: how is a business process
implemented? (number of tasks in a job, relative
size of tasks, nature of tasks, degree of customisation), and
b. the behaviour view: when is a business process
executed? (sequencing of tasks, task consolidation, scheduling of jobs, etc.);
 the participants in the business process considering
a. the organisation structure (elements: roles,
users, groups, departments, etc.), and
b. the organisation population (individuals: agents
which can have tasks assigned for execution and
relationships between them);
 the information the business process uses or creates;
 the technology the business process uses, and
finally;
 the external environment other than the customers.
In this paper we aim at investigating if this framework is relevant and valid. At first, we were quite

2. Initial BPR framework and set of best practices

2.1. The BPR framework
We have explored in the literature several
frameworks and business process analysis models
that were potentially suitable for business process
redesign. In [4] we explain how we have derived an
extended framework for implementing BPR best
practices. It is derived as a synthesis of the WCA
framework by Alter [3], the MOBILE workflow
model by Jablonski and Bussler [5], the CIMOSA
enterprise modelling views of Berio and Vernadat [6]
and the process description classes of Seidmann and

Fig. 1. Framework for BPR implementation.

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

459

Table 1
BPR best practices classified according to our BPR implementation framework
Framework elements

Best practice name

Definition

Customers

Control relocation
Contact reduction
Integration

Move controls towards the customer

Reduce the number of contacts with customers and third parties
Consider the integration with a business process of the customer or
a supplier

Products

None

Operation view

Order types
Task elimination
Order-based work
Triage

Task composition
Behavioural view

Resequencing
Knock-out
Parallelism
Exception

External environment

Trusted party
Outsourcing
Interfacing

Organisation: structure

Order assignment
Flexible assignment
Centralisation
Split responsibilities
Customer teams
Numerical involvement
Case manager

Determine whether tasks are related to the same type of order and,
if necessary, distinguish new business processes
Eliminate unnecessary tasks from a business
Consider removing batch-processing and periodic activities from a
business process
Consider the division of a general task into two or more alternative
tasks or consider the integration of two or more alternative tasks
into one general task
Combine small tasks into composite tasks and divide large tasks into
workable smaller tasks
Move tasks to more appropriate places
Order knockout decisions in a decreasing order of effort and in an
increasing order of termination probability
Consider whether tasks may be executed in parallel
Design business processes for typical orders and isolate exceptional
orders from normal flow
Instead of determining information oneself, use results of a
trusted party
Consider outsourcing a business process in whole or parts of it
Consider a standardised interface with customers and partners
Let workers perform as many steps as possible for single orders
Assign resources in such a way that maximal flexibility is preserved
for the near future
Treat geographically dispersed resources as if they are centralised
Avoid assignment of task responsibilities to people from
different functional units
Consider assigning teams out of different departmental workers that
will take care of the complete handling of specific sorts of orders
Minimise the number of departments, groups and persons involved
in a business process
Appoint one person as responsible for the handling of each type of
order, the case manager

Organisation: population

Extra resources
Specialist-generalist
Empower

If capacity is not sufficient, consider increasing the number of resources

Consider to make resources more specialised or more generalist
Give workers most of the decision-making authority
and reduce middle management

Information

Control addition

Check the completeness and correctness of incoming materials and

check the output before it is sent to customers
Instead of requesting information from an external source, buffer it by
subscribing to updates

Buffering
Technology

Task automation
Integral technology

Consider automating tasks

Try to elevate physical constraints in a business process by applying
new technology

460

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

confident, as it is a synthesis of previously validated

frameworks described in the literature. However, we
had concerns about whether all the elements should be
dealt with during the implementation of BPR. Also,
we wanted to know how important each element is in
the implementation process and whether practitioners
give some aspects of the framework a higher priority
over others.

been in close contact with. The intention of these case

studies is to get a firmer idea of the importance of the
various best practices and the usability of the
framework. We discuss the subsequently raised issues
around this framework in Section 3.
Next, we have conducted an electronic survey
amongst experienced Dutch and UK practitioners in
BPR to assess the validity of our framework against
their practices. These results are discussed in Section 4.

2.2. BPR best practices

Within our framework for BPR implementation, we
have gathered and classified best practices in BPR.
Over the last 20 years, best practices have been
collected and applied in various areas, such as business
planning, healthcare, manufacturing, and the software
development process (e.g. [810]). In [4] we have
described such best practices intended to support the
redesigner of a business process in facing the technical
BPR challenge: the implementation of an improved
process design. The paper also discussed the
advantages and drawbacks of each best practice. A
qualitative evaluation was also undertaken to assess
the best practices against their impact on time,
flexibility, quality and cost issues. The Devils
quadrangle of Brand and Van der Kolk [11] was used
for the purpose.
Table 1 summarises the identified best practices
within the BPR implementation framework and their
definitions. For more information, references, case
studies and examples of these practices, the interested
reader is referred to the previously mentioned
reference [4].
Several authors mentioned the above best practices
as having a positive impact on enhanced business
processes. However, many questions remain unanswered: how far are these rules used in redesigning
real processes within organisations? Is it possible to
classify them? Can we derive a top 10 list of best
practices? Is our qualitative assessment of the best
practices impact valid? etc.
2.3. Approach
To answer some of the identified questions in this
section, we set out the following approach. As a start,
we used our framework in implementing some best
practices in BPR to two Dutch companies we have

3. Framework and best practices validation

through two case studies
3.1. Case studies description
In 2002, we applied the BPR framework and its best
practices within two Dutch organisations. The occasion for cooperation with these organisations was their
participation in a longitudinal study into the effectiveness of workflow management (WfM) systems in
which the authors are involved [12]. These types of
systems enable the automatic coordination of business
processes (for more information, refer to [1,5]).
The first organisation involved is a local municipality of 90,000 citizens in the northern part of the
Netherlands. The specific department in question was
the Urban Management Service, responsible for
sanitation, parking facilities, green spaces, and city
districts. This service employs over 300 civil servants.
The second organisation is a provincial office of the
national public works department, employing about
1000 civil servants. This offices primary responsibility is the construction and maintenance of the road
and water infrastructure in its province.
Both organisations selected their invoice processing workflow to be supported by a WfM system. The
municipality handled about 10,000 invoices in the
years 2000 and 2001. For the public works office, that
amount was approximately twice as high. In addition
to these processes, the municipality also aimed to
automate its purchasing process. On a yearly basis,
the Urban Management Service puts out some 700
purchasing orders.
For both organisations, we estimated the effects of
applying BPR to the three mentioned processes and
implementing WfM technologies to support these
processes. The identified list of BPR best practices

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

played a major role in this estimation, as explained in

the next paragraph. We have created new computer
simulation models on the basis of the models
representing the initial situation in both companies.
The development of the various simulation models for
each business process under the various circumstances
is depicted in Fig. 2 and explained in further detail
below.
A number of alternative models were built for each
of the two invoicing processes and for the purchasing
process. The starting point for all these models is the
initial model. This initial simulation model captures
the existing process structure and includes real
behavioural data of the process on the arrival patterns
of cases, resource capacity, service time characteristics, etc. In [12] we explain in some more detail how
this initial model is built, how the data has been
gathered and how the model is validated against the
real world.
The next type of model that was built models the
situation where only the Integral technology best
practice was applied to the current process, i.e. the use
of WfM technology to support the existing process.
We will refer to this type of model as the WfM-model.
The typical effects of WfM technology on process
performance were incorporated, as identified in [1]
and [5]. For example, transportation activities that
exist in the current process were eliminated from the
initial model, because WfM technology will take care
of these.
Next, in preparation of the BPR-model, we
determined for each single best practice whether it
would be applicable in the context of the specific
process. If so, we constructed for such a single best

461

practice a simulation model where the effect of this

specific best practice was incorporated on top of the
existing WfM-model. To model the best practice
accurately, we used estimations from experts from
both organisations to approximate the local effect of
such a single BPR best practice within the process. In
addition, we used the results from simulating such a
WfM-model extended with just this single best
practice to determine the overall performance
improvement.
Finally, for the three business processes under
consideration, the BPR-model incorporates a subset of
all applicable best practices. This subset seemed the
best combination in terms of performance improvement. Whether a best practice was included in the final
BPR-model was determined during two workshops
that involved end users, managers, and IT professionals for both organisations.
3.2. BPR implementation and results
The simulation model (BPR-model) was used to
test how appropriate it was to apply other BPR rules
than the implementation of WfM technology alone.
The evaluation of the relevance of the various best
practices is given in Table 1. The table displays the
results of the evaluation of each best practice for the
three processes. The table indicates whether a best
practice was applicable and, if so, whether it produced
significant results and whether it was subsequently
included in the finally redesigned process.
It is important to note here that the municipality
was primarily interested in reducing labour cost by
reducing the average service time spent on cases for

Fig. 2. Derivation of simulation models.

462

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

Table 2
Evaluation of best practices applicability to studied business processes
Framework element

Best practice

Municipality

Public works

Invoices

Customers

Operation view

Behavioural view

Organisation: structure

Organisation: population

Control relocation
Contact reduction
Integration
Order types
Task elimination
Order-based work
Triage
Task composition
Resequencing
Knock-out
Parallelism
Exception
Order assignment
Flexible assignment
Centralisation
Split responsibilities
Customer teams
Numerical involvement
Case manager
Extra resources
Specialist-generalist
Empower
Control addition

Information

Buffering

Technology

Task automation
Integral technology

External environment

Purchasing

Invoices

U
U





U

U**

U**

U**


U**

U
U

U

U**

U
U

U

U**

U*

U**

U**

U
U



U**


U


U
U






U


U**

U**

Trusted party
Outsourcing
Interfacing

A: applicable, S: significant, I: incorporated in the redesign, U: yes, : no.

*
Significant using a two-sided 90% confidence interval.
**
Significant using a two-sided 99% confidence interval.

both the invoice processing and purchasing process,

while the public works office aimed at reducing the
cases average lead time for their purchase order
process. Significant results were therefore measured in
these terms.
Analysing Table 2, it follows that the best practices
with the widest applicability are: Task elimination,
Task composition, Integration, and Specialistgeneralist. These best practices were applicable in
all three cases. The Task elimination and Task
composition best practices were estimated to be the

most effective, as they delivered significant results in

respectively all and two out of three processes.
Statistical significance was established when there was
no overlap between the two-sided 99% confidence
intervals of the performance measure in question (i.e.
lead time or labour cost), resulting from simulation
experiments of the WfM model and the same WfM
model extended with the best practice in question. The
relative importance of the various best practices is
given in Fig. 3. This score is determined by calculating
how often the best practice was considered to be

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

463

Fig. 3. Best practices relevance for the BPR-model.

applicable and significant for the process and whether

it was finally incorporated in the redesign. The highest
score was obtained by the Task elimination best
practice. The latter was incorporated in all three
redesigned processes.
A relative comparison between the different
situations is graphically displayed for each of the
three processes in Fig. 4.
From the results in Fig. 4, it can be seen that the
BPR-model in all situations by far delivers the best
results. In the case of the public works invoicing
process, the average lead time was expected to be
reduced by almost 80%, where the sole introduction of
WfM technology would only account for 25%. It is
perhaps not very surprising that the public works
office decided to combine the implementation of WfM
technology with a redesign of their invoice processing
(i.e. to apply the selected other best practices as well).
The actual process performance 3 months after the
implementation was completed, as could be determined on basis of the management information the
running WfM system generated, concords almost
exactly with the estimated effects.
The municipality, however, did not decide to
conduct BPR, which they found too great an organisa-

tional risk considering the expected gains. Currently,

the implementation of the WfM systems has been put
on hold within this organisation, because of budget
problems and the municipality is looking for a cheaper
alternative for the WfM technology.
3.3. Implications for the framework and the set of
best practices
Using this framework during the analysis and
redesign stages proved to be useful to identify and
structure the involved business processes. On the one
hand, it helped us to distinguish the proper angles for
applying the best practices to the current process. On
the other hand, it stimulated the discussions during the
workshop with the domain experts. All the elements of
the framework were dealt with and none of them
appeared to be irrelevant for the redesign. As far as the
best practices are concerned, the implementation and
interviews conducted with both the managers and
actors of the processes revealed that a smaller set of
best practices could have been considered. This is
reflected in Fig. 3, as a number of best practices
emerges as having a potentially much higher impact
on business processes than others. As a consequence,

464

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

Fig. 4. Relative reduction in performance indicators.

we have established a top 10 list of most popular

best practices. Our ranking was established using the
two case studies where the 10 best practices appeared
to be most popular. To classify the best practices that
had equal/close ranking, we used our literature review
(i.e. how often was a given best practice cited in the
literature?). The list of the 10 most popular best
practices is provided in Table 3. The table also links
each best practice to an element of our framework.
At this stage of our research, we have managed to
answer some of our initial questions about the best

practices: how far are these practices used in

companies? Is it possible to classify them? Can we
derive a top 10 list of best practices? However, we
still felt that the framework and an assessment of how
useful the best practices could be, would be accurate
only if they are applied to a wide range of companies
or if they are validated by trusted and experienced
practitioners in BPR. So, we decided to conduct a
survey amongst Dutch and UK practitioners, as the
first option is only feasible on the long term.

Table 3
Most popular best practices in business process redesign

4. Framework and best practices validation using

a Dutch/UK survey

Ranking

Best practice

Framework element

1
2
3
4
5
6
7
8
9
10

Task elimination
Task composition
Integral Technology
Empower
Order assignment
Resequencing
Specialist-generalist
Integration
Parallelism
Numerical involvement

Operation view
Operation view
Technology
Organisation: population
Organisation: structure
Behavioural view
Organisation: population
Customers
Behavioural view
Organisation: structure

The survey took place in 2003 and targeted wellestablished practitioners in the BPR field. To select
potential participants to our survey, we decided to
focus on Dutch and UK practitioners as we were based
in both countries and wanted to exploit our local
contacts with BPR practitioners. The survey excluded
pure academics or, to be more precise, academics who
could not show evidence of experience in BPR
projects within/with real organisations. Practitioners

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

465

Table 4
Participants profile
Practitioners

Dutch sample

UK sample

Sample size
Response rate (%)
% of BPR practitioners
Years of experience
Self-expertise assessment

31
42
92
Range: 735; average: 14.8; mode: 15
Range: 510; average: 7.8; mode: 8

60
20
92
Range: 1035; average: 20; mode: 10
Range: 410; average: 6.75; mode: 6

were selected according to the company they

represented (e.g. well-established consulting groups)
and also according to their track record in BPR, e.g. as
could be concluded from published case studies they
authored. The survey was conducted using an online
questionnaire that was sent to participants using
emails (mapping in the questionnaire). The email
explained the context of the survey. The context was
emphasised by pointing the participants to an online
review of previous BPR surveys. The review aimed at
positioning our survey against the previous ones and
clarifying our expectations. Table 4 shows that for
both samples (Dutch and UK) the vast majority of
practitioners had more than 15 years of experience and
ranked their own expertise in the field close to 7 on a
scale of 110.
Many surveys exist in the literature about business
process reengineering [13,14,1619]. However, we
could not find specific ones related to business process
redesign, i.e. surveys related to how the process should
be articulated in terms of tasks and resources, for
example, as opposed to how to manage the changes in
an organisation. The aim of our survey is to validate
our framework (refer to Fig. 1) and the impact of the
10 selected most popular best practices (refer to
Table 3). As a consequence, we have decided to test
the following hypothesis.
Table 5
Practitioners ranking of framework elements
Framework elements (N = 25)

Average

Mode

Median

Customer
Information
Product
Operation view
Behaviour view
Organisation
Technology

3.72
3.36
3.40
3.08
2.92
2.88
2.84

4
4
4
3
3
3
3

4
4
4
3
3
3
3

H1. The framework for BPR implementation that we

have developed is valid and covers all possible aspects
practitioners look for when redesigning processes.
H2. All the rules that have been identified as best
practices (refer to Table 3) are indeed applied extensively by practitioners.
4.1. Survey structure
The survey consisted of four major parts. The first
part included general questions to determine our
respondents profile and to assess their expertise in
BPR. The results are indicated in Table 4. The second
part included questions designed to validate our
framework. We have asked the participants to rate and
discuss how much and how often they focus on each
frameworks element when undertaking a BPR
project. (mapping into questionnaire) To avoid
confusion on the meaning of each framework element,
we did not use its designated terms but explained it.
The term Operations view of the framework might
be ambiguous and interpreted differently by participants. So, instead we used The way a workflow
operation is implemented (i.e. the number of tasks in a
job, the relative size of tasks, the nature of the tasks,
the degree of customisation). The results of the
second part are indicated in Table 5 and in Fig. 5. The
third part of the survey listed the 10 most popular best
practices we have initially selected. Participants were
asked to express whether they had used any of the best
practices listed and, if so, how often. (mapping into
questionnaire) Again, to avoid confusion on the
meaning of the best practices, we phrased our question
in a way that would avoid ambiguity. For example, for
the Task Composition best practice rather than
asking if the participants had already used the best
practices we asked: Have you combined small tasks

466

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

Fig. 5. Comparing initial and validated BPR framework.

into composite tasks and divided large tasks into

workable smaller tasks?. In addition, for each best
practice we provided a link to a simple and short
example where the best practice is applied. For the
earlier mentioned best practice we provided the
example of a conference registration procedure. The
results are indicated in Table 6.
Participants also ranked the impact of each best
practice on the quality, the flexibility, the time and the
cost performances of a given best practice (not
covered in this paper). At the start of the survey, these
performance criteria were explained. For each best

practice when the respondent was asked for his or her

opinion on its impact, hyperlinks were provided to the
original explanation of these notions. In the last part,
participants were asked to indicate whether they had
used the best practices in their most successful (and
least successful) project. They were also asked to
indicate the best practices that contributed the most to
a BPR projects success. In this paper, we only relate
the results of the first three parts.
Finally, it is important to note that the questionnaire
was tested before it was sent out on a group of 10
academics, both experts and non-experts in BPR. The

Table 6
Classification and level of usage of best practices amongst participating practitioners
Ranking

Best practice

% Usage

Frequency

Framework element

1
2
3
4
5
6
7
8
9
10

Task elimination
Integral business technology
Task composition
Parallelism
Specialist-generalist
Resequencing
Integration
Empower
Numerical involvement
Order assignment

94
94
89
88
88
88
76
76
76
53

All participants used it six times or more

16/17 participants used it between two and five times
15/17 participants used it between two and five times
15/17 participants used it between two and five times
15/17 participants used it between two and five times
15/17 participants used it between two and five times
13/17 participants used it between two and five times
13/17 participants used it between two and five times
13/17 participants used it between two and five times
9/17 of those who used it did so two and five times.
The remaining used it only once

Operation view
Technology
Operation view
Behavioural view
Organisation: population
Behavioural view
Customers
Organisation: population
Organisation: structure
Organisation: structure

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

feedback we gathered led us to make several

adjustments, especially with respect to its length.
The comments we received gave us the impression
that there was no confusion on the meaning of the
questions. That is why we felt quite confident that
these concepts would be clear to a community of
practitioners in BPR, carefully selected for their broad
experience in the field.

4.2. Framework analysis and validation

Practitioners from UK and the Netherlands were
asked to assess the validity of our BPR framework.
For each element of the framework (Customer,
Information, Product, Operation and Behaviour view
of a business process, Organisation and Technology)
participants were asked to express how much they
focused on each element when redesigning a business
process. They were offered to answer never,
sometimes, often or almost always. We have
ranked these options from 1 to 4. Table 5 displays the
results of participants answers using the latter
ranking. In this table we use a point estimate for
the means (average values) for the scores practitioners attributed to each element of the framework as
an indication of their distribution. We provide
median and modes as the major indicators for the
scores.
At this point in the survey, our sample size is
composed of 25 participants. We acknowledge that a
sample size of at least 30 would have been better for
the analysis (see e.g. [21]). However, we remain quite
confident in the interpretations we can derive from
this analysis because of the heterogeneity of the
population. The respondents we targeted were carefully selected amongst trustful and experienced
practitioners in BPR with years of hands-on experience. We analysed whether there was a bias to an
industry over the other, but this was not the case. Also,
we compared our sample size and response rate to the
other BPR surveys we know of. Although our sample
size at this point of the survey is relatively low, the
response rate of 27% is higher than that of most other
studies. Finally, our qualitative interpretation of the
results, taking into account the numerous comments
we received from the respondents, ensures a valid
interpretation.

467

In our initial framework, we included the eight

aspects mentioned in the table above. Aspects with the
highest scores were customer, information and
product. All of these gathered an average, mode and
median score expressing to their use as often or
almost always. From these, customer and product
are the least surprising ones as they centre on what is
being produced for whom. Less obvious is the
information aspect. Our interpretation is that the
aspect refers to an area where large improvements can
be made using BPR. This is supported by some of the
comments of the respondents, e.g. that of respondent
NL22: Invariably, process improvement requires
information improvement (be it not always computerised). The comment also makes the distinction
clear between the information and technology
aspect.
Before the survey, we expected that all of the eight
frameworks elements would receive fairly high
scores. However, the aspects behaviour view,
organisation and technology received an average
score which indicated a use lower than often (i.e. 3).
However, considering the modes and medians of these
aspects, most of them still reach a level indicating their
use as often. To us, this seems a positive and much
more meaningful indication of these aspects relevance, considering the qualitative scale used.
In the case of the organisation aspect, the
qualitative support was not so evident from the few
respondents who considered this aspect to be valuable.
However, when the respondents were asked in the
following part of the survey for missing parts from our
framework, we received the following responses of
elements that were used almost always:
1. To ensure the solution fits with the culture of the
business. (UK13)
2. The design of the business process is also affected
by the competences of the people that execute
them. (NL21)
Both of these seem to clearly refer to the organisational aspect of a BPR initiative, which makes it
difficult to decide that this aspect is not relevant within
a BPR initiative.
Finally, we received one more response of an aspect
that was used almost always, although it was not
present in our original framework:

468

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

3. Lack of process information is another origin of

problems. (NL28)
In our view, this comment is relevant as it extends
our interpretation of the information aspect from
the information the business process uses or creates
to the information the business process uses or
creates and information about the process execution
itself.
Our overall conclusion is that our framework is
recognised and supported by the community of BPR
practitioners. Aspects of the framework that seem to
be in the centre of attention in a BPR initiative on both
sides are customers, products and information.
These are presented in bold in the figure of the
validated framework, to signify their relevance.
There is only one problematic aspect, i.e. the
organisation aspect. There is no immediate and firm
support for incorporating it within our framework.
However, our inquiries into missing elements seem to
suggest that this aspect is relevant nonetheless. We
have graphically depicted the questionable status of
this aspect by showing it in dotted lines in our
validated framework (see Fig. 5). One explanation for
the ambiguous status of the organisation aspect
could be that BPR initiatives typically try to break
away from existing organisational structures and
attitudes, exactly to achieve great benefits. As
respondent NL30 puts it: Considering the organisation aspect in a BPR initiative is only the last step, only
to be taken after logical and executable process
models have been designed.
To summarise, this section addressed the validation of our first hypothesis (i.e. the framework for
BPR implementation that we have developed is valid
and covers all possible aspects practitioners look for
when redesigning processes). The discussion in this
section and the comparative initial and final framework (refer to Fig. 5) indicates that the framework
elements as we have identified them from analysing
the literature and applying to the case studies
described in this paper are all relevant and should
be addressed in a BPR effort. However, we realised
that some adjustments had to be made to our initial
framework, essentially as to the relative importance
of the elements. The organisation element was the
main questionable part of the framework, as discussed earlier. Also, the survey revealed the lesser

importance of the behaviour and information

elements. Not surprisingly, the customer and
product elements appeared to be crucial in the
redesign effort.
4.3. Best practices analysis and validation
As far as the best practices are concerned, we
wanted to validate through this survey the validity of
our classification of top 10 best practices (refer to
Table 3). For this sake, participants were asked
whether they have used a specific best practice in their
BPR projects and, if so, how often (refer to results in
Table 6). At this stage 30% of the participants stopped
the survey, only providing answers about the framework validation. We think that this is linked to the
relative length of the survey.
In the following, we indicate percentages of
participants who recognised a best practice as valuable
and how often they have used it in their projects.
Obviously, the figures in Table 6 support our initial
classification of best practices as largely popular
amongst practitioners. For each best practice, the
majority of participants mentioned to have used them
at least 25 times in earlier BPR projects.
Interestingly, although most participants agreed
that they would mostly focus on the customer, the
product and the information elements of our
framework when redesigning a business process, the
widely applied rules are chosen and classified
according to the operation, the technology and
the behavioural elements of our framework. We
might conclude that in order to obtain a business
process of which the aims are customers oriented
(good service, good product, good information
flow), process designers need to focus primarily
on the operational and behavioural views of a
business process, as well as on the structure of the
processes.
Finally, it is noticeable that the bottom of the list
includes the Order assignment, the Numerical
involvement and the Empower rules (all related
to the organisation element of our framework). This
is consistent with the validated framework, as
participants seemed not to focus much on the
organisational part when redesigning a business
process. We had reflected this fact by displaying the
Organisation element in a dashed box in Fig. 5.

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

Some clues to support this exclusion might be found in

some participants comments about the relevant best
practices. For example, on the Order assignment best
practice (Let workers perform as many steps as
possible for single orders) a participant (UK8) noted
that he never uses the rule because the segregation of
duties may limit the stages that one operative can
perform as may the limit of an individual employees
training. Simply having one operative do more of the
process is not necessarily an improvement. On the
Numerical involvement (Minimise the number of
departments, groups and persons involved in a
business process) another participant (UK9) claims
that Au contraire, we recognised that core processes
cuts across department and invite the group to work
together. Finally, on the Empower best practice
(Give workers most of the decision-making authority
and reduce middle management) the same participant
justified the non-usage of the rule by claiming that
This involves redefining the organisation structure
and governance authorities.
As far as validation of the second hypothesis is
concerned (i.e. all the rules that have been identified as
best practices (refer to Table 3) are indeed applied
extensively by practitioners), our initial ranking of
best practices (Table 3) is close to the findings of this
surveys participants (Table 6). In both cases, the first
three mostly used best practices are Task composition, Integral business technology and Task
elimination. These three rules are strongly related
to the essence of BPR as IT is reported to be an
essential component of reengineering and involves
analysing tasks usefulness [20]. Besides, in both
tables both the Numerical involvement and the
Integration rules are not that popular. We have
already explained that this is consistent with the
validated framework. Finally, compared to Table 3,
one important difference is the Parallelism best
practices position. Parallelism implies Considering
whether tasks may be executed in parallel. The
participants assessment of this best practice is, we
think, much more in-line with the potential benefits it
might bring (drastic cut of process time). Possibly, the
parallelism rule simply was not relevant to the specific
projects we have discussed in Section 3. This confirms
the usefulness of conducting a survey on top of
assessing rules applicability to a small number of real
organisations.

469

5. Discussion
A valid question now is: How do our framework
and the set of best practices support BPR practice? Our
ideal of delivering a comprehensive and sound
methodology for BPR practitioners definitely requires
us to take some further steps, as we will reflect upon in
our conclusion. However, some guidelines can already
be given. First, the conducted case studies and the
survey allow us to indicate which areas are vital during
a redesign process. We demonstrated that the focus
must go beyond the process in itself and must embrace
the customers, the product, the information views and,
to a lesser extent, the technology and the organisation
part. In [15] there is support for this direction, as is
demonstrated through a set of industry case studies:
reengineering is more successful when a wider view of
BPR is adopted.
Next, we have classified in this paper the best
practices according to the mentioned framework
elements. We strongly suggest that our previous paper
[4] should be read in conjunction with the present one.
That previous paper provides a comprehensive review
about each best practice, in particular the main
advantages and disadvantages of using each of them
are discussed. The review also includes a checklist of
the currently available tools and techniques that
support each best practice. It also indicates, qualitatively, the impact of the best practice within the devils
quadrangle framework. In other words, it gives general
indications for the kind of effects one may expect from
applying a best practice. Finally, it provides pointers to
case studies where the best practice was applied
successfully.
In lack of a mature methodology we feel we can
propose ourselves right now, we recommend practitioners to [22] and [23]. First of all, these provide
cook book like procedures to e.g. identify, describe,
and analyse current business processes. Secondly, they
describe how workshops can be used to stimulate
creativity among its participants to come up with
improved and organisationally accepted designs. We
believe that our validated framework could be used to
structure such a workshop-centred approach, ensuring
that each of its elements receives attention. Furthermore, the best practices we describe may provide
guidance to the workshop participants for the kind of
changes they can suggest. For example, a question a

470

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

workshop facilitator may want to ask on the basis of

the Task Composition best practice is: Are there
any steps in the existing process that could be
combined, so that the process could be executed faster
and chances on hand-off mistakes are minimised?.
In summary, we provided through this and previous
work insight into the advantages and disadvantages of
each best practice, the context in which they should be
applied (i.e. the frameworks element), and which
best practices are most popular amongst practitioners
and should therefore be considered with priority by
redesigners.

6. Conclusion: BPR framework and best practices

validation
In this paper we have described a framework for
classifying best practices in business process redesign.
The purpose of this framework is to set up the grounds
for a directed methodology for BPR projects implementation. The framework indicates the major areas a
practitioner needs to focus on when redesigning a
project: the customer, the products and the information flow. It also points out to other important areas
such as the behavioural and the operation view of a
process and the technology that should support the
redesigned business process.
We provided a set of redesign rules that we believed
were popular amongst practitioners. The rules are
gathered to provide practitioners ideas on different
implementation options for a new process. In this
paper we have demonstrated the validity of the BPR
framework and best practices through two steps. In the
first step we have tested them on two organisations that
considered a BPR project. In the latter step, we have
used a survey amongst practitioners in BPR. Both steps
allowed us to conclude on a validated framework (refer
to Fig. 5 and to hypothesis H1 in Section 4) and to
select a top 10 list of best practices amongst practitioners (refer to Table 6 and hypothesis H2 in Section
4). The future research directions are as follows:
 At first, we would like to analyse the impact of the
top 10 best practices on the flexibility, the cost, the
time and the quality as perceived by practitioners.
 Secondly, to investigate for all best practices when,
where and how to apply or not apply them. This

means giving indications to the size of the business

process or the tasks involved. Also, it should study
the relative impact of best practices on a business
process. In this area, in [24] the popular combination of the empower and the triage best practices are
studied (leading to decentralisation and task
consolidation). It is proved, using mathematical
models, that this combination is sub-optimal in
many cases.
 At last, to provide users with a methodology in
applying best practices. This includes using our
classification of the best practices within the
framework for BPR implementation as a basis
(which was one of the purposes of this paper), and
deriving a guideline to the order/conditions in
which the best practices should be implemented.
Several authors have already published relevant
work in this area. In [25] a streamlining of business
process redesign rules is provided. In [26] a
framework is provided for analysing BPR in
conjunction with several strategic dimensions.
However, these approaches do not consider the
full set of best practices we described in this paper.

Acknowledgement
Part of this work by Hajo Reijers was done at the
Process Management Research Center (Babson
College) during a sabbatical leave. Part of this work
by Selma Liman Mansar was done at London
Metropolitan University while she was working there.

References
[1] W.M.P. van der Aalst, A.H.M. ter Hofstede, M. Weske, Business process management: a survey, in: Proceedings of the
2003 International Conference on Business Process Management (BPM 2003), Lecture Notes in Computer Science, vol.
2678, Springer-Verlag, Berlin, 2003, pp. 112.
[2] F. Leymann, D. Roller, M.T. Schmidt, Web services and
business process management, IBM Systems Journal 41 (2)
(2002) 198211.
[3] S. Alter, Information Systems: A Management Perspective,
Addison Wesley, Amsterdam, 1999.
[4] H.A. Reijers, S. Limam Mansar, Best practices in business
process redesign: an overview and qualitative evaluation of
successful redesign heuristics, Omega, The International Journal of Management Science (available on-line).

S.L. Mansar, H.A. Reijers / Computers in Industry 56 (2005) 457471

[5] S. Jablonski, C. Bussler, Workflow Management: Modeling
Concepts, Architecture and Implementation, International
Thomson Computer Press, London, 1996.
[6] G. Berio, F. Vernadat, Enterprise modeling with CIMOSA:
functional and organizational aspects, Production Planning &
Control 12 (2) (2001) 128136.
[7] A. Seidmann, A. Sundararajan, The effects of task and
information asymmetry on business process redesign, International Journal of Production Economics 50 (2/3) (1997)
117128.
[8] J. Martin, The Best Practice of Business, John Martin Publishing, London, 1978.
[9] P. Butler, A strategic framework for health promotion in
Darebin, A Report to the East Preston and Northcote Community Health Centers by the Center for Development and
Innovation in Health, Center for Development and Innovation
in Health, Melbourne, 1996.
[10] J. Golovin, Achieving Stretch Goals: Best Practices in Manufacturing for the New Millennium, Prentice-Hall, New York,
1997.
[11] N. Brand, H. Van der Kolk, Workflow Analysis and Design,
Kluwer Bedrijfswetenschappen, Deventer, 1995 (in Dutch).
[12] H.A. Reijers, Performance improvement by workflow management systems: preliminary results from an empirical study, in:
Proceedings of the Sixth International Conference on Enterprise Information Systems (ICEIS 2004), vol. 3, INSTICC,
Porto, 2004, pp. 359366.
[13] M. Zairi, D. Sinclair, Business process re-engineering
and process management: a survey of current practice
and future trends in integrated management, Business Process Re-engineering & Management Journal 1 (1) (1995)
830.
[14] T. Guimaraes, W. Bond, Empirically assessing the impact of
BPR on manufacturing firms, International Journal of Operations & Production Management 16 (8) (1996) 528.
[15] D. Grant, A wider view of BPR, Communications of the ACM
45 (1) (2002) 8590.
[16] J. Valimaki, T. Tissari, Risk management focus in business
reengineering initiatives, in: Proceedings of the IPMA Symposium on Project Management: Managing Risks in Projects,
E & FN Spon, London, 1997, pp. 233242.
[17] P. ONeill, S.A. Sohal, Business process reengineering: application and successan Australian study, International Journal
of Operations & Production Management 18 (9/10) (1998)
832864.
[18] T.J. Crowe, P.M. Fong, T.A. Bauman, J.L. Zayas-Castro,
Quantitative risk level estimation of business process reengineering efforts, Business Process Management Journal 8 (5)
(2002) 490511.

471

[19] R.S. Maull, D.R. Tranfield, W. Maull, Factors characterising the

maturity of BPR programmes, International Journal of Operations and Production Management 23 (6) (2003) 596624.
[20] M. Hammer, J. Champy, Reengineering the Corporation: A
Manifesto for Business Revolution, Revised paperback
edition, Harper Business, New York, 1994.
[21] J.T. Roscoe, Fundamental Research Statistics for the Behavioural Sciences, 2nd ed. Holt, Reinhart & Winston, New
York, 1975.
[22] G.A. Rummler, A.P. Brache, Improving Performance: How to
Manage the White Space on the Organization Chart, 2nd ed.
Jossey-Bass, San Francisco, 1995.
[23] A. Sharp, P. McDermott, Workflow Modelling: Tools for
Process Improvement and Application Development, Artech
House, London, 2001.
[24] A. Seidmann, A. Sundararajan, Competing in informationintensive services: analyzing the impact of task consolidation
and employee empowerment, Journal of Management Information Systems 4 (2) (1997) 3356.
[25] H.J. Harrington, Business Process Improvement: The Breakthrough Strategy for Total Quality, Productivity, and
Competitiveness, McGraw-Hill, New York, 1991.
[26] W.J. Kettinger, T.C. Teng, Aligning BPR to strategy: a framework for analysis, Long Range Planning 31 (1) (1998)
93107.
Selma Limam Mansar is an assistant
professor in Management Information
Systems at the college of business
sciences at Zayed university. She worked
at the London Metropolitan University
and Eindhoven University of Technology.
She obtained her Ph.D. in industrial engineering in 1999 at the National Polytechnic Institute of Grenoble. Her current
research interests are business process
redesign, simulation, business-to-business systems and knowledge
management applications.
Hajo A. Reijers is an assistant professor
in Information Systems at the Department
of Technology Management of Eindhoven University of Technology. Before
this, he has worked for several management consultancy firms, most recently as
a manager with Deloitte. His main interests are business process redesign, information systems, workflow management
systems, Petri nets, and simulation.