INCREASING THE VALUE OF PROCESS MODELLING
John Krogstie
IDI, NTNU, Sem Sælandsvei 7-9 7030 Trondheim, Norway
Vibeke Dalberg, Siri Moe Jensen
DNV,Veritasveien 1, 1322 Høvik, Norway
Keywords: Business process modelling and re-engineering.
Abstract: This paper presents an approach to increase the value gained from enterprise modelling activities in an
organisation, both on a project and on an organisational level. The main objective of the approach is to
facilitate awareness of, communication about, and coordination of modelling initiatives between
stakeholders and within and across projects, over time. The first version of the approach as a normative
process model is presented and discussed in the context of case projects and activities, and we conclude that
although work remains both on sophistication of the approach and on validation of its general applicability
and value, our results so far show that it addresses recognised challenges in a useful way.
1 INTRODUCTION
Enterprises have a long history as functional
organisations. The introduction of machinery in the
18th century lead to the principle of work
specialisation and the division of labour, and on to
the need of capturing, structuring, storing and
distributing information and knowledge on both the
product and the work or business process. Business
process models have always provided a means to
structure the enormous amount of information
needed in many business processes (Hammer, 1990).
The availability of computers provided more
flexibility in information handling, and led to the
adoption of modelling languages originally
developed for systems modelling like IDEF0 (IDEF-
0, 1993). The modelling of work processes,
organisational structures and infrastructure as an
approach to organisational and software
development and documentation is becoming an
established practice in many companies. Process
modelling is not done for one specific objective
only, which partly explains the great diversity of
approaches found in literature and practice. Five
main categories for process modelling are proposed
based on Curtis, Kellner, and Over (1992), Totland
(1997), and Vernadat (1996):
1. Human-sense making and communication to
make sense of aspects of an enterprise and to
communicate with other people
2. Computer-assisted analysis to gain knowledge
about the enterprise through simulation or
deduction.
3. Business Process Management
4. Model deployment and activation to integrate
the model in an information system
5. Using the model as a context for a system
development project, without being directly
implemented (as it is in category 4).
In an ongoing project on model-based network
collaboration, we have investigated the practice and
experience of process modelling across four
business areas and a number of projects and
initiatives in a large, international company. Our
objective was to identify possible improvements and
facilitate potential sharing of relevant resources,
aiming towards an optimisation of value gained from
modelling and models. Merriam-Webster Online
defines value as: “something (as a principle or
quality) intrinsically valuable or desirable”. We have
aimed for a company-wide, inclusive scope in our
use of the term value, guided by what has been
deemed relevant by involved stakeholders.
Three important observations were made during
the early stages of the project:
70
Krogstie J., Dalberg V. and Moe Jensen S. (2006).
INCREASING THE VALUE OF PROCESS MODELLING.
In Proceedings of the Eighth International Conference on Enterprise Information Systems - ISAS, pages 70-77
DOI: 10.5220/0002457800700077
Copyright
c
SciTePress
Even within projects a variety of objectives was
found, spanning the categories presented above. A
corresponding variety was found in tools, methods
and attitudes to the potential value of modelling.
In some initiatives there were significant
divergence of expectations to the modelling results
and value - between different stakeholders and also
over time.
Communication and sharing of resources
between projects were mainly done through more or
less ad-hoc reuse of models and personnel
personally known by project workers in advance.
From this we made three assumptions:
Single project value and stakeholder
satisfaction could be increased by to a larger degree
focusing on, communicating and prioritizing
between diverging expectations and objectives.
This would require a common platform for
communication about modelling initiatives
expectations, objectives, and other attributes.
Such a platform could also facilitate reuse of
relevant knowledge, tools, models, methods and
processes between units and projects.
These assumptions lead to the development of a first
version of a framework proposal on best practice for
increasing the value of process modelling and
models. This proposal consists of a taxonomy, a
recommended model of activities for process
modelling value increasing initiatives, and links to
relevant knowledge and best practices for each step
of the process. Work leading up to this work has
been reported in (Dalberg et al, 2003; Dalberg et al
2005; Krogstie et al, 2004; Krogstie et al, 2005).
The rest of this paper presents the methods used
in our work, from identification of needs,
development and assessment. We then give an
overview of our first version of the framework of
best practice for increasing the value of process
modelling and models, and discuss its applicability
with regard to challenges identified in earlier
projects. Finally, we conclude on the applicability
and usefulness within the limitations of our
validation, and indicate needs for further
development of the framework as well as for more
large-scale validation within a wider scope.
2 RESEARCH METHODS
The research presented in this paper is based on
qualitative analysis of a limited number of case
studies. According to Benbasat, Goldstein, and
Mead (1987), a case study is an approach well suited
when the context of investigation takes place over
time, is a complex process involving multiple actors,
and is influenced by events that happen
unexpectedly. Our situation satisfies these criteria,
and the work has taken place within the frames of a
three year project, including one in-depth case study,
and several other less extensive studies. In deciding
whether to use case studies or not, Yin (1994) states
that a single case study is relevant when the goal is
to identify new and previously not researched issues.
When the intent is to build and test a theory, a
multiple case study should be designed. The
intention of our study has been to find out how to
increase the value of modelling and models in an
organisation. There has not been reported much
research within this area earlier, and we have
therefore chosen a multiple case approach for the
work presented in this paper, in order to investigate
this research area closer.
The framework for increasing value of process
modelling and models presented in this paper has
been developed through an iterative process, refining
the model. So far we have been through four
iterations.
In the first iteration we studied the modelling
initiative in a particular project in detail, using
observation, participation, and semi-structured
interviews. After initial explorative research, we
focused on identifying the expectations and
experiences towards the modelling and the models,
on their score related to process modelling success
factors, as well the extensive reuse of the models
across the organisation, viewing this as possible
knowledge creation and sharing as a part of
organisational learning. An initial hypothesis on
process modelling value was established, based on
our findings regarding the importance of the relation
to the context of modelling versus the context of use.
In the second iteration, we went through semi-
structured interviews with representatives of several
different modelling initiatives throughout the
organisation to survey their experience with
modelling, especially with respect to benefits and
value of reusing knowledge through models across
projects and organisation. A number of initiatives
were selected for the study where we were able to
get in-depth knowledge from those involved in the
process. An interview guide for interviews with key
stakeholders was established. These interviews were
focused on expected and experienced use and value
from the modelling efforts in the case study, aiming
at identifying as many expectations as possible,
including any that may not have been documented in
project documentation, because they were not
considered directly relevant for the project goal.
After initial open questions, the interviews were
structured around keywords from the work of
Sedera, Rosemann, and Doebli (2003) concerning
INCREASING THE VALUE OF PROCESS MODELLING
71
“process modelling success”. Documentation of the
study is based on these interviews, studies of project
documentation and models. The information from
the interviews was partly structured through the use
of the interview guides. The guides were used as
basis for structuring contact summary sheets with
the main concepts, themes, issues and questions
relating to the contact (Miles and Huberman, 1994).
As a third iteration we carried out a workshop
with a group of modelling experts, discussing the
framework in relation to their own experiences
through numerous process modelling projects. This
resulted in an updated version of the framework.
In what has so far been our last iteration, we
included the framework in an actual business project
using action research, where one of our researchers
also acted as a modeller. This was an informal test of
the framework, but gave valuable input to updating
it. We also saw the value of the framework in a
modelling initiative through this test, where it gave
positive guidance for the modelling. The next
iteration of the development of the best practice
framework should be to conduct more formal tests.
Our results and approach this far has certain
limitations relative to internal validity (Miles and
Huberman, 1994), as representatives of some of the
involved roles have been followed more closely than
others. As for descriptive validity (what happened in
specific situations) the close day to day interaction
with the users, especially in the first and the last
iteration by one of the researchers, give us
confidence in the results on this point. As for the
interpretive validity (what it means to the people
involved) we have again in-depth accounts from
central people in main roles, but again not all the
involved roles have been represented to the same
degree. The same can be said on evaluative validity
(judgements of the worth and value of actions and
meaning). That we find many results that fit the
categories of existing theoretical frameworks gives
us confidence on the theoretical validity of the
results.
3 A FRAMEWORK FOR
INCREASING THE VALUE OF
PROCESS MODELLING
This best practice framework aims to increase the
value of the modelling and models through enhanced
awareness about current and future stakeholders, any
(potential) conflicts of interest, stakeholder
expectations and potential value to be gained, as
well as any negative effects increasing total cost.
Based on this knowledge, decisions regarding
resource allocation, modelling methods and tools,
responsibilities etc can be made to optimize the
value of a modelling activity and its resulting
models, on a project level as well as on an
organisational level. The basic elements of the
framework are a recommended main process (see
Figure 1) and some basic concepts, elaborated on in
the description of each step in the main process.
Context is the surroundings of an initiative that
might influence decisions. Value is identified in
relation to the identified context, but also on
potential value outside the initial project scope. The
practice focuses on the strategies and practice
around the modelling and the models.
The recommended process is initiated when a
need for modelling has been identified. Its three
main steps are detailed below.
3.1 Identifying Context
Identifying the context is mostly about expressing
the circumstances of the identified need for
modelling, as a basis for further communication,
prioritization and planning. It will usually coincide
with the writing of an application for funding,
development of a project mandate and/or a project
plan. At this step one should keep within the scope
of the initial need, usually expressed in traditional
project documentation with formal obligations. The
main issues to be clarified are detailed in Figure 2,
and include:
Identification of the context of the modelling or
model activity/initiative, including users and
other stakeholders, uses, and objectives.
Identification of the organisations installed
base, including existing reusable models or
descriptions and other relevant tacit or explicit
constraints.
Figure 1: The overall framework.
ICEIS 2006 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION
72
Figure 2: Identifying context.
There are different actors related to a modelling
initiative and a model, holding one or more roles.
Users are using the models or participating
personally in the modelling in order to achieve
objectives. Other stakeholders may not be using the
models directly, but extract value from planned
objectives. Techniques e.g. from user-centred design
is useful at this stage in the identification of
stakeholder types. Use includes how the modelling
and models are going to be used in order to achieve
the objectives. Objectives are the goals and purposes
of the modelling and models. Installed base includes
tacit and explicit assets already existing in the
organisation that will have influence on the
modelling and model context. Constraints include
issues such as personal and organisational
knowledge, which may be tacit or explicitly
expressed constraints, organisational guidelines or
instructions (explicit constraints), existing tools and
languages etc. Reusable models are models or other
documentation that were created for other purposes,
but that could be reused in the new project.
3.2 Identifying Potential Value
In step 1, we identified the context where the
modelling and the models were meant to play a role.
In step 2, “Identify potential value”, the aim is to
capture any (potential) extra and positive benefits of
the modelling and models, exceeding the primary
objectives captured in step 1. Value may be
connected to the resulting models, or to the
modelling activity in itself.
Often the objectives identified in step 1 will
relate to the modelling or model initiative, while any
potential value to the rest of the organisation will
typically be ignored in the formal project
documentation developed at this stage – due to a
lack of awareness, or to avoid complicating
responsibilities and bindings.
Value can be explicit and easy to grasp, but also
tacit. Tacit value, e.g. the improved understanding of
a work process for a modeller originally producing
models for others, are often not explicitly captured
in traditional project documentation, but may still
affect decisions before or during a project, or the
perceived value of the project in retrospect. Future
reuse of the models can be an added value of the
current modelling and models, especially if this
potential is taken into account at an early stage.
Figure 3: Identifying potential value.
INCREASING THE VALUE OF PROCESS MODELLING
73
Figure 4: Choosing practice.
3.3 Choosing Practice
The choice of a suitable practice should be based on
the identified contexts of the modelling and models,
as well as the identified expected value. Modelling
practice include reuse strategy, methods, languages
and tools, while managing practice define how to
manage the modelling, the models and the work
processes. The general framework of quality of
models and modelling languages inspired by
organizational semiotics (Krogstie and Sølvberg,
2003) is especially helpful here relative to modelling
practice related to methods, languages, and tools,
having the stakeholders of the models and the goals
of modelling already defined. When goals or
stakeholder types are changed during a modelling
project, one needs to reassess these aspects, and
potentially select a new modelling language, method
or tool.
Sense-making versus corporate memory
We have chosen to differentiate between
modelling for sense-making and for corporate
memory. These concepts can be helpful for
expressing fundamental differences in expectations
to a modelling initiative, often rooted in personal
worldviews emerging as strong opinions on
modelling use and approaches. Totland (1997)
addresses modelling for sense-making and corporate
memory, and the relation to objectivistic and
constructivistic worldviews.
The corporate memory models are reflecting the
organisation, and will exist as a reference point over
time. The sense-making models are used within an
activity in order to make sense of something in an
ad-hoc manner, and will usually not be maintained
afterwards. Sense-making and corporate memory
can be seen as the two endpoints of a scale, where
you have examples of mixed types of models in
between.
These concepts express and explain one type of
differences and disagreements between stakeholders,
drifting within projects, or conflicting approaches in
modelling activities that would otherwise be
expected to have much in common.
The choice of the formality of the modelling
practice should be based on the previously identified
contexts, and where these fit on the line with sense-
making and corporate memory as the two extremes.
Sense-making initiatives generally require a low
level formality of practice.e When the context is
corporate memory, a more formal approach is
needed. The choice of methods, tools and languages,
as well as the choice of managing practice should
reflect the level of formality needed. High formality
requires more managing than low formality.
Table 1: Comparing modelling for sense-making and
corporate memory.
Sense-making Corporate memory
The modelling process is
the goal
The model itself is the
goal
The actual use is often
documented
The intended use is
often documented
Collects the natural
structures
Collects the formal
structures
Identified people
important
General user-roles
important
Less formal methods,
tools and languages
Formal methods, tools
and languages
Roles not important, more
ad-hoc
Roles important.
Often used only for a
specific activity or project
Often re-use across the
organisation
The models are “thrown
away” after use
The models are stored
and re-used
Management of the work
process, models and
modelling not important
Management of the
work process, models
and modelling
important
When identifying the context of the modelling
activity, the optimal position on the sense-making –
corporate memory axis is crucial in order to be able
ICEIS 2006 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION
74
to choose appropriate methods, languages and tools,
as well as formality for the managing practice.
4 APPLYING THE FRAMEWORK
During our research we have studied and
documented several cases throughout the
organisation. Through this we have identified
expected and experienced value of modelling work
and models, as well as experienced challenges. In
this chapter we quote some of the reported
(potential) value. We will then look into how the
framework addresses the reported challenges.
4.1 Identifying Potential Value
The stakeholders in our case studies indicated many
valuable outputs in addition to those initially
intended from modelling initiatives and the use of
models. Some of these are:
Communication:
The high-level models encouraged an
agreement among the management participants
that was vital for the rest of the project, creating
important common references, identification
and enthusiasm.
The models triggered communication, being
something that everyone could relate to.
“Three boxes and some arrows: This is a
fantastic communication tool”.
Communication was initiated and facilitated by
and through the models.
The models help the participants understand.
Learning:
The modelling process itself turned out to be a
learning experience for the participating domain
experts, increasing their knowledge about the
processes.
Through the workshop sessions the participants
learned a lot from interacting with each other,
“new” information was uncovered, and
understanding improved.
People understand themselves better after a
modelling session.
The participation in the modelling process of
domain experts is important. The result would
not have been the same if modellers from
outside created the models based on interviews.
The models helped taking care of and storing
the competence of people in the organisation.
Modelling is seen as a mechanism to extract
knowledge from people’s heads.
Training takes less time when process models
were used.
Long-term benefits:
The process model gives the organisation one
language and one tool for everyone in the
organisation; a common frame of reference.
Simple and effective diagrams show what is
important for the organisation.
Through modelling AsIs, and not only ToBe,
best practise is secured and not forgotten.
The models are used in marketing towards
potential customers.
There is a marketing value in telling the world
that they have documented processes.
5 CHALLENGES OF
MODELLING AND MODELS
In order to extract more value from the modelling
initiatives and the models, we will in the following
address some of the major identified challenges in
our case studies, and examine how the framework
could solve or indicate a solution to these. For each
paragraph we state the challenge, then how it is
addressed in the framework.
Challenge 1: To keep the models and other
descriptions updated and consistent
Example: It becomes difficult to keep the
models updated as the complexity increase, and the
number of non-integrated tools increases.
Framework application: The framework
suggests careful analysis of the expected model
context before choosing the modelling practice.
Considering the future complexity when choosing
methods, language and tools will make model
management easier. The framework also states the
importance of viewing the management of the
models as a specific activity, stressing the
importance of appointing a model responsible. This
is a different role than the modelling responsible or
the work process responsible (process owner).
Challenge 2: The models are used in situations
they were not intended for.
Example: Models are often created primarily for
one objective. This is challenging when others want
to use them as basis for other work, especially if the
original assumptions are not documented.
Framework application: Through an analysis in
the early phase of the modelling activity, identify the
primary use as well as potential future use and
additional potential value. Accommodation of
indications of future use of the models should be
INCREASING THE VALUE OF PROCESS MODELLING
75
considered when choosing the modelling and the
managing practice.
When in a re-use situation, where a modelling
initiative is going to re-use earlier developed
models, it is important to investigate the context the
models were created for, and what modelling and
managing practice have been used. The decision of a
re-use strategy should be based on this investigation.
Challenge 3: To handle situations when the
modelling starts out as an informal activity, but the
resulting models develop into a process defining
tool. The original language and tools often do not
meet new expectations for the model to be kept
updated, be scaleable, and extendable with new
functionality. The experience is that the chosen tool
and language often do not fit into this new scenario.
Framework application: Awareness of where on
the scale of sense-making versus corporate memory
the models were initially created, and where on the
scale the models have ended up (and where they can
be expected to end up). Sense-making models do not
require a very high level of formality, while
corporate memory models often do. Being conscious
about this will make it easier to identify what has to
be changed in the modelling and managing practice
in order to align with the new situation.
Challenge 4: To produce views of the model
according to different needs and users.
Example: Not being able to produce views of the
models adapted to the specific user and the objective
of the use creates challenges. Specific users and
specific objectives of use require adapted views of
the model. The creation of these is a challenge, both
technically and as regards content.
Framework application: Identify the users and
other stakeholders as parts of the context, analyse
their background knowledge and needs, and what
each of them are going to use the models for.
Methods, language and tools should then be chosen
based on this.
Challenge 5: The models often restrict and limit
the communication.
Example: High level models are easy to agree
upon, but real gaps between the model and current
situation stay uncovered. A model is only one view
of the world. When a model is the communication
generating artefact, the discussions often leave out
those issues not included in the model.
Framework application: Carefully identify the
context and the potential value of the modelling and
models before creating the models. Consciousness
about how to increase the potential value of
communication will potentially help creating a more
fitting model. Awareness of the limitations of a
model and its restrictions is the key.
Challenge 6: To implement the models in the
organisation, particularly outside the modelling
team.
Example: It is a challenge to make the models an
integrated part of the organisation, and to involve the
users to the extent that they feel an ownership and
responsibility for them. When the person doing the
modelling leaves the project and the modelling is
left to the domain experts to finish, implement and
keep updated, experience shows that the focus on the
models often fades. If the modeller leaves too early,
the models may not be implemented.
Framework application: Identify all the expected
users and other stakeholders during the initial phase
of the modelling activity, look into their expected
areas of use and identify potential value. By
choosing a modelling practice to increase the value
across all identified stakeholders, ownership and
usefulness is improved even for stakeholders not
participating in the modelling. If many stakeholders
should be involved in the modelling one can use
techniques such as ”modelling conferences”
(Gjersvik et al 2004)
Challenge 7: To be conscious about distributing
the responsibility of the modelling, models and
processes correctly.
Example: One person was responsible for
everything that had to do with the processes and the
models.
Framework application: The framework makes
distinctions between the activities of managing the
modelling, the models, and the work processes. One
role is related to the management of the modelling,
another to the management of the models, a third to
the management of the work processes.
Challenge 8: During organisational changes,
models may have to be merged as processes are
unified. Different modelling tools and languages
increase the challenge.
Example: Several as-is processes were to be
harmonized and their documenting models merged
into one common process model. The models were
created for different user groups, originated in
different organisational units and also countries. The
modelling processes were also different, involving
different types of people.
Framework application: Such models are most
likely based on different methods, languages and
tools, created for different objectives, uses and users
and other stakeholders. The historic context and the
modelling and managing practice of each of the
models should be investigated in order to establish a
re-use strategy and choose the correct current
modelling and managing practice.
ICEIS 2006 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION
76
6 CONCLUSION AND FURTHER
WORK
Based on extensive research across units and
projects in an international company, we have
identified expectations, challenges and experience
pointing to potential increase in value from
modelling activities. To support the realization of
these values, a Modelling Value Framework has
been developed.
The Value Framework has been evaluated against
challenges and experiences of earlier modelling
initiatives, as well as tested in a modelling project.
There are clear indications that further development
and use of the framework will facilitate
communication and alignment within and between
project initiatives and organisational units, thus
potentially increasing value from projects through
improved relevance and quality of results as well as
reduced cost.
Our research has been practically oriented, aiming
towards identification of the important issues in real-
life modelling projects and activities, both with
regard to the actors’ motivation and their experience.
Based on the broad investigations we have made, we
are confident that our results are valid for the case
company.
We expect our findings to be reproducible for other
enterprises of similar size and complexity, but this
still remains to be shown.
Even within the presented enterprise, on a practical
level, there is still a way to go to implement and
collect real-life experience with the framework. Our
studies demonstrate feasibility and advantages of
use, but do not address the actual adoption of the
framework by practitioners not involved in the
development.
We have identified advantages both on a project and
organisational level, and we expect that the project
level advantages will be sufficient to motivate for
the use of the framework – and that the
organisational level advantages can be realized this
way. This assumption however still has to be tested
– and a successful implementation in the whole
organisation will, as a minimum, require a dedicated
dissemination and marketing effort.
REFERENCES
Benbasat, I., Goldstein, D. K. and Mead, M. (1987) "The
case research strategy in studies of informations
systems" MIS Quarterly (11:3) p 369-386
Curtis, B., Kellner, M., Over, J. "Process Modelling,"
Communication of the ACM, (35:9), September 1992,
pp. 75-90.
Dalberg, V., Jensen, S. M., Krogstie, J. Modelling for
organisational knowledge creation and sharing, in
NOKOBIT 2003. Oslo, Norway
Dalberg, V., Jensen, S. M., Krogstie, J. Increasing the
Value of Process Modelling and Models, in
NOKOBIT 2005. Oslo, Norway
Gjersvik, R., J. Krogstie, and A. Følstad, Participatory
Development of Enterprise Process Models, in
Information Modeling Methods and Methodologies, J.
Krogstie, K. Siau, and T. Halpin, Editors. 2004, Idea
Group Publishers.
IDEF-0: Federal Information Processing Standards
Publication 183(1993) Announcing the Standard for
Integration Definition For Function Modelling.
Hammer, M. Reengineering Work, Don’t automate,
Obliterate. Harvard Business Review, 1990
Miles, M. B, and Huberman, A. M. Qualitative Data
Analysis, SAGE Publications 1994
Krogstie, J. and A. Sølvberg, Information systems
engineering - Conceptual modeling in a quality
perspective. 2003, Trondheim, Norway:
Kompendiumforlaget.
Krogstie, J., V. Dalberg, and S.M. Jensen. Harmonising
Business Processes of Collaborative Networked
Organisations Using Process Modelling. in
PROVE'04. 2004. Toulouse, France.
Krogstie, J., Dalberg, V., Jensen, S. M., Using a Model
Quality Framework for Requirements Specification of
an Enterprise Modeling Language, in Advanced
Topics in Database Research, volume 4, Siau K.
Editor. 2005, Idea Group Publishers.
Sedera, W., Rosemann, M. and Doebeli, G. (2003) “A
Process Modelling Success Model: Insights From A
Case Study”. 11th European Conference on
Information Systems, Naples, Italy
Totland, T. (1997). Enterprise Modelling as a means to
support human sense-making and communication in
organizations. IDI. Trondheim, NTNU.
Yin. R. Case study Research. SAGE Publications. 1994
Vernadat, F. (1996) Enterprise Modelling and Integration.
Chapman and Hall.
INCREASING THE VALUE OF PROCESS MODELLING
77