KNOWLEDGE BASED PROCESS REENGINEERING
Jiří
Voříšek
Department Information Technologies
Prague University of Economics
W. Churchilla 4
130 67 Praha 3
e_mail: vorisek@vse.cz,
http://nb.vse.cz/~vorisek/
Published: March 2000
Key words: individual and
organisational knowledge, business processes, strategic management, management
of IS/IT
In the late 1990s changes in economic and
social environment have accelerated. The main accelerators are new
communication and information technologies and their applications - e.g.
E-commerce, E-learning, virtual teams and virtual organisations, customer
relationship management, and supply chain management. Many authors (e.g.
Drucker, 1999, Hesselbein, 1999, Thurow, 1997) are documenting these in their
papers very thoroughly. Examples of changes typical for our global information
society are as follows:
·
Increase in the importance of knowledge for the
achievements of individual, enterprise, even a nation (together with poor
visions for little qualified),
·
Increase in the importance of alliances and
logistics chains,
·
Mergers of formerly separated branches,
·
Increase in the stressing of process
organisation of business and its high flexibility,
·
Increasing influence of value system
(enterprise or national culture) on managing processes,
·
Extremely fast development of information and
communication technologies that influences all areas of human activities.
All
these changes are also reflected in new requirements on business management and
business information system. Therefore the concepts and principles of strategic
management and systems integration are to change, too. The imperative of the
moment is interdisciplinary and multidimensional approach towards the majority
of problems that are being solved.
This
article points out the necessity and possibility of integration of business
processes with three types of knowledge: a) individual knowledge of individual
employee, b) together with knowledge accumulated in the business over
time and c) together with knowledge and competencies of business partners.
We are going to introduce a hypothetical
example in order to make the explanation of general principles easier. The
majority of the principles are going to be demonstrated on the example. The
example is to demonstrate following:
·
Changes in style of thinking that are required
by the new economic environment
·
Changes in strategic management of processes
and knowledge
·
The important role of alliances and outsourcing
when ensuring fast respond to market changes
Ø Small garden-tillage with a store
Ø So far none IS that would be used by the company, IS for accounting and
wages are outsourced
Ø SWOT:
·
(-) limited number of customers
·
(-) seasonality of sales (green-houses are not
used since last year - expensive heating)
·
(-) domestic flowers only
·
(+) skilled arrangers of bouquets
·
(+) creative management (owner)
·
(+) new business opportunities arising from new
information and communication technologies
Ø The primary target of the company is to multiply its turnover and its
market share within one year, the increase in profit is just a secondary target
(the target is more the value of the company than its immediate profit)
Ø Taking the advantage of maximum existing opportunities - co-operation, incorporation into chains,
modern methods of payment, information and communication technologies, ...
Ø New business model:
·
Orders via the Internet
·
Involvement of a customer in creating the
bouquet (types of flowers, design of bouquet, packing, selection of card going
with the bouquet, place and time of delivery)
·
Prepared bouquets for less creative customers
·
Purchase of flowers from other countries
(especially from Dutch wholesalers)
·
Narrowing the core business during the year,
liquidation of the garden-tillage and concentrating on arranging and deliveries
of single flowers or complex flower decorations to end customers
·
Outsourcing of the shipping (Post, FedEx,...)
Ø The company concentrates on its core business, still outsources IS for
accounting and wages. New functions of IS are also going to be solved by
outsourcing (forming alliances with prominent providers of information services)
in 4 consecutive projects (knowledge and competencies of business partners will
be utilised):
1)
web orders - completely outsourced
through web services provider
§
provider
interface: (a) output - catalogue and flowers price list,
(b) input - customer's order (by fax in the first phase, then by electronic means - email)
§
service
provider price: 5% of the value of the order (motivation for
web provider)
§
new
costs: approximately 1 000 USD (PC with internet
access) and telephone charges
§
realisation
period: 4-6 weeks
§
garden-tillage
is going to be run for the duration of one more season
§
deliveries
to remote customers - cash on delivery
2)
system for communication with bank -
acceptance of credit card payments and usage of SET
3)
system for communication with
flower wholesaler
4)
system for communication with FedEx - deliveries to foreign countries
As we have already mentioned in (Voříšek,
1999) there is a possibility of
describing the interdisciplinary and multidimensional concept of systems
integration by four basic concepts. The elementary concept is called "Processes, areas and sources of
business management". This concept places IS/IT among other enterprise
sources and processes and it shows their mutual links - see Fig. 3.
Any area can be divided into five levels (these
levels are modifications of organisational processes levels proposed by Scheer,
1999) according to the processes that exist in the areas of management.
First
Level - strategic management - sets strategic
characteristics of every area and their links. The
main outcomes of this level are answers to these questions:
Ø
What are the company's goals and priorities,
Ø
What products/services is the company going to
provide and who are its customers,
Ø
What alliances and co-operative relationships
is the company going to form, what competencies is the company going to apply
in co-operation and which one does it expect its partners to perform (Voříšek, 1998),
Ø
How are the material and information flows
going to run among partners in chain - rough framework,
Ø
What human, knowledge/information and financial
sources are going to be necessary for the company and how are they going to be
obtained (how are the excessive sources going to be released),
Ø
What metrics are going to be used to measure
the degree of fulfilling the goals.
The above mentioned example is describing this
level. Management of the company took the advantage of new IS/IT possibilities
(communication with customers, suppliers and banks via Internet) together with
the knowledge and competencies of other firms (web provider, bank FedEx) in
order to redefine its core business and to create completely new range of
products and services and to find new target audience.
The
second level is Process Design. This level defines or modifies business processes that lead to
production of products and services defined in level one. The target of this
level is to propose the business processes in such a way that they would lead
to production of products/services with defined quality, in minimum time with
the consumption of the least possible enterprise sources. Main outcomes of this
level are:
Ø
Design of main processes (processes leading to
customer),
Ø
Design of supporting processes,
Ø
Description of each process (initiating event,
individual activities and their flow, job positions responsible for individual
activities etc.),
Ø
Time calculation of duration of typical process
instance and calculation dealing with typical consumption of sources (material,
finance, human capital, etc.),
Ø
Setting of metrics that are going to be
observed during the process, their dimensions and frequencies of their observing.
The aim is to initiate a feedback (the arrow in the Fig. 3) that would allow
continuous improvement of the process.
One of the key trends in this area is maximum self-service of customers
and employees through IS/IT. An example could be the wholesale communication
with its suppliers. It does no longer order the goods with its suppliers, but
the suppliers themselves have to access wholesaler's IS and have to control the
deliveries in such a manner that there is always optimal stock of given goods
available.
Getting
back to our example, the main processes could be defined as follows:
1)
Creation of the catalogue and the price list of flowers and other
services
2)
Acceptance and execution of
order for one bouquet,
3)
Acceptance and execution of complex order (flower decoration) - in this
and the above mentioned process the self-service type of ordering is used ,
and these supporting processes:
1)
Purchase of flowers from Dutch wholesaler,
2)
Growing own stock of flowers etc.
Following
metrics can be assigned for these processes (data in brackets quote the
dimensions in which the metrics are observed):
·
Orders
taken by web provider (number, price, day, month, type of order, type of
customer, residence of customer),
·
Total
costs of orders (day, week, month, type of order, type of customer),
·
Profitability
of individual products (type of
product, territory),
·
Time
needed to respond to the order (week, month, type of order),
·
Number
of claims (day, month, type of order, type of customer) etc.
The
third level is the Process Control. Level two was dealing with processes on a general level, it created a
process model. All instances of processes are going to be run on the base of
this model. As this model states estimated duration of individual instance of
process and estimated consumption of business sources during a typical process,
it sets basic data for detailed planning and control of processes and business
capacities.
The design of processes on level two is an
occasional activity (created model of process can remain unchanged over several
years). On the other hand, control of processes is a continuous activity and
its outcomes are:
Ø
Detailed operative plan of processes and
consumption of business sources. This plan may show what business sources are
going to be activated, the order of execution of accepted orders, place of
execution of an individual order of an individual customer etc. Its aim is to
satisfy requirements of all customers, if possible at all, without running into
bottleneck, respectively having idle business sources,
Ø
Regulatory intervention into running of
processes.
The importance of the operational level of
management in the era of E-commerce has substantially grown. The reason for
this is the fact that due to this business form it is even harder to estimate
the volumes of orders and the capacities needed for executing these. The
operational level is therefore responsible for the optimal sizing and optimal
allocation of business capacities (human capital, technology, finance)
according to changes in the volume of orders.
In our example of a florist store planning of
processes and capacities could include: planning of shifts, recruitment of
part-time employees during high season, price adjustments of products and
services according to changing number of orders etc.
Process Workflow is occurring on the last fifth level. The processes are running according to
the process model defined by level two and by detailed plan set by level tree. Outcomes of this level are
already individual products and services.
Creation
and deliveries of ordered flowers and complex flower decorations will be
happening on this level in our example.
Above this operative level, there is the fourth level called Process Monitoring. The results of metrics value analysis are
then in relevant aggregation passed to upper levels:
Ø
To level 3 - for better process control
Ø
To level 2 - for optimisation of processes
models
Ø
To level 1 - for evaluation and changes of
business strategy.
Note: Activities run on levels 1-4 are managerial processes that can be described
and analysed by the same methods and tools as operational processes of level 5.
There are very tight links among individual
areas of management. For instance, the main process "execution of an
order" from the area business logistics must be bound to company's
organisation (specific job positions are responsible for individual activities
of the process). This process must be financially secured and supported by
IS/IT. Processes from the three remaining areas (human resources, finance,
IS/IT) can be either supporting processes to the main processes, or can exist
directly as a part of main processes (such as subprocess - order of bouquets
through web).
Systems integration must secure not only the
integration of processes of every individual area of management, but also
integration of processes from all areas.
A good example could be the integration of new
business processes with business organisational structure in transition from
the commodity or geographically oriented organisational structure to organisational
structure depending on language and national culture.
Let us have a look now what questions must be
solved while integrating IS/IT with individual levels of managerial and
operational processes. We are going to use and further develop the proposal of
Scheer (Scheer, 1999). The difference between our and Scheer's approach is
based on Scheer's usage of another specific level so-called Process
Application. This level includes software support of individual operational
processes. On the other hand, our concept assigns relevant data and software
sources to every of the five above-mentioned levels - further explanation is
given in the following table.
|
Application (software
sources) |
Information Sources |
Strategic Manage-ment |
· EIS (Executive Information System), BI (Business Intelligence) · Datawarehouse |
· Information about market development · Long-term statistics of company results according to several
dimensions · Benchmarking · Information about products and services of potential co-operators · Information about national and company culture |
Processes Design |
· Modelling tools (process diagrams, stage diagrams,...) · Simulation tools · Tools for quality management |
· Reference models of products · Reference models of processes · Statistics from running the processes in the past |
Process and Capacity Control |
· Control systems for time and capacity management · Tools for project management |
· Status and plans of processes · Status and plans of business sources |
Monitoring |
· Monitoring systems · Systems for data analysis · It is often included in applications for operational processes support
|
· Sensors measuring operational processes and organisational sources · Data of individual metrics |
Process Workflow |
· Workflow tools · Modules of standard ERP systems (often include also applications of
higher levels) |
· Detailed data describing current status of individual instances of
processes and individual enterprise sources |
One of the key problems of the above-described
concept is the process design and optimisation. BPR has become one of the most
frequently discussed topics of professional conferences and discussions of
business managers in the 1990s. However, its results (as well as the results of
systems integration projects) are not always satisfactory.
The following approach used for process design
and optimisation was developed by the team of the Information Technologies
Department. It was also verified in an outsourcing project of a large utility
company. The base of the concept is the criticism of the mechanical approaches
to design and reengineering of processes which are too technically oriented, do
not respect the knowledge of people participating in the processes and do not
allow human creativity in the process.
The result of the classical process design and
process reengineering (especially if ARIS Toolset type of tools is used) is
detailed design of process that contains exactly defined network of activities.
For every partial activity in the process following data are described: its
inputs and outputs (sometimes also algorithm of an activity), responsible job
position for given activity, maximum duration of an activity, maximum costs per
activity, conditions necessary for running the activity, etc.
This detailed design of process offers many
advantages (e.g. when describing processes of car production):
Ø Every running of the process and its outcome are identical with the
ideal values that were counted with in the design of the process,
Ø Running of the process has well predictable duration and costs,
Ø The activities of a process can be carried out by relatively unqualified
(but well trained) labour force,
Ø Specialists designing the model are very creative; on the other hand,
creativity during running the process could lead to very negative results such
as disturbing the optimal flow of activities, low quality output etc.
Above-described approach may, however, bring
number of disadvantages to other situations:
Ø As soon as the running of the process depends on many conditions,
external factors and their combinations, then the model of the process gets
very complicated. Imperfect description that may not be solvable due to some
combinations of conditions can lead to abnormal termination of the process,
Ø There are certain situations were the optimal model might not even be
agreed upon. For instance, when describing the process of development of
information strategy we may be able to describe the optimal structure of
document of information strategy. However, very likely we are not going to be
able to describe the whole process of development of the information strategy
in the detail that was mentioned above. We could try to do so, but this
detailed proposal would be probably rejected by the specialists developing the
strategy as they would soon reveal the shortages of the design for the given
situation,
Ø The strict definition of the process does not allow to take advantage of
new ideas and proposals that could be brought up by the employees taking part
in the running of the process.
Due to the reasons described above, Knowledge Based Process Reengineering distinguishes
four levels of modelling and optimising a process. The difference among
levels is based on different level of knowledge and experience required from
the employees participating in running the process. It is up to the creators of
the design to choose the optimal level for each of the processes. Individual
levels of description are characterised as follows:
(1)
The first level - the least detailed design of the process - describes
the process through these characteristics:
·
Target of the process (product/service and its
characteristics)
·
Event activating the given process
·
Role (job position responsible for the whole
process)
·
Qualitative and quantitative metrics of the
process and
·
Limiting conditions of the process (financial
or time limit for running the process).
If we used this level to
define the process of "Creating the catalogue of offered flowers",
then the process could be defined, for example, this way:
· Targets of the process: to
create a catalogue for individual market segments that is responsive to
national cultures and to pricing policies in these territories
· Events activating given
process: change in customers values, change in pricing policies in the
territories, new collection created by designers
· Job position responsible for
the process: marketing director
· Qualitative and quantitative
metrics of the process: number of offered flowers in the catalogue, number of
pricing policies etc.
· Limiting conditions: the
catalogue has to be changed at least four times a year; the offer of prepared
bouquets has to be at least in three pricing policies, calculated profit from 1
bouquet must exceed 10%
Every
other level of process design always keeps the characteristics of the previous
one and adds some more.
(2)
The second level is extended by defining
the outcome of the process (required structure of the final document, model of
a product, photography of a sample etc.).
If we used this level to define the process of
"Creating the catalogue of offered flowers", then we would add one
more binding definition of catalogue structure for individual market segments.
Note: we are not going to give examples for the two
following levels due to their high amount of text requirements; Attentive
readers would be surely able to fill them in by themselves anyway.
(3)
The third level is extended by:
·
List of activities that are parts of the
process (however, the flow of the activities is not defined),
·
List of
roles, job positions taking part in the process (roles are not assigned to
activities)
·
List of all external inputs into the process
(inputs are not assigned to activities)
·
List of all appropriate software tools to
support the process.
(4)
The fourth level is extended by:
·
Flow of activities,
·
Inputs and outputs of every activity,
·
Responsible roles are assigned to individual
activities.
The process described here is called the
knowledge based process reengineering as it distinguishes two types of
knowledge - accumulated enterprise knowledge (saved in the company's
database, in existing definition of company processes and standards etc.) and empoyees'
knowledge (those employees engaged in the running the process).
The first level of describing the process is
used in the case when the accumulated enterprise knowledge is not sufficient,
e.g. when running of every process is very different due to changed conditions
and external factors. This situation may happen especially in the processes of
strategic management. This is the reason why the first level requires highly
qualified and creative employees who are going to be able to plan and implement
all so far not defined characteristics of the process during running the process.
For
defining the process of "Creating the catalogue of offered flowers",
it would be advisable to use the first level at the beginnings of the firm's
pursuing the new business strategy. The
reason for this is that the firm is lacking the experience with optimal
structure of catalogue for web pages and for different territories. At this
stage, it is also important to hire a prominent marketing specialist who would
create the catalogue.
On the other hand, the fourth level of
describing the process assumes highly qualified and experienced specialists who
design the process. Therefore consultant firms are very often being hired to
help in designing the process, as these people know the "best
practice" of given area of interest. The fourth level at the same moment
represents the highest level of standardisation of running the process.
It is important to realise that the more we are
approaching the fourth level, the higher the costs of process definition are
and the lower the costs of process running including the costs of managing the
process are. The reason for this is that when realising a process that is
defined at the fourth level, it is not necessary for the employees to figure
out the best way of workflow. This task has already been fulfilled by the designer
of a process. This also leads to the conclusion that the higher the number of
process instances, the more appropriate a detailed process definition.
KBPR suggests that the level of definition of
the process may change over time. As the knowledge of employees taking part in
the process realisation increases, there is growing opportunity to further
develop the design of the process. The more detailed the design gets, the
easier it is to handle the work over to less qualified employees. This de facto
leads to accumulation of enterprise knowledge. Gradual accumulation of
knowledge makes the company less dependant on concrete people and their
knowledge. We can therefore state that the accumulation of enterprise knowledge
is one of very important supporting processes - as e.g. maturity model of
company's development (CMM, 1995). All this rises a very interesting
question - how to motivate the employees to hand over their knowledge?
Accumulation of enterprise knowledge by
specifying the model of business processes is in a way (from the systems point
of view) an analogy to evolution development of layered architecture of system
(Voříšek, 1997). The development of understanding and automation of certain
area of reality usually undertakes following steps:
(1)
At the beginning all the problems of area of reality are solved by
intuition, without any in advance elaborated plan
(2)
Gradual understanding leads to elaboration of various methods of problem
solving. The application of the methods requires creative approach in every
step
(3)
Further deepening of understanding the area and experience with methods
leads to elaboration of consensual approach to solving. The solution is divided into steps with exactly defined consequences,
each step has exactly defined inputs and outcomes. The activities in every step
are, however, defined only approximately
(4)
By further understanding is the consensual approach changing into the algorithm of solution. Algorithm is already in detail
defined solution of the problem that there is no more necessity of any creative
approach of people and therefore computers are in charge. Algorithm in this
stages becomes a part of individual application software;
(5)
Gradual standardisation of formerly different algorithms of solving
similar problems leads to transmission of algorithm to lower level - into a standard application package, such
as ERP;
(6)
The algorithms that have been standardised the most and that are crucial
to the efficiency of the system become part of hardware.
The tools that are appropriate for automation
of process support depend also on the level of process design. The more we are
approaching the fourth level, the more appropriate are the standard application
packages such as ERP (SAP R/3, BAAN etc.). On the contrary, the more we are
approaching the first level, the more appropriate are universal tools such as
MS Excel, respectively different applications of EIS.
There is no necessity in praxis to stay with
the four levels of description of the process as defined here. When proposing a
concrete business process, it is always at first necessary to evaluate the
level of employees' knowledge and the state of accumulated enterprise
knowledge. Then, on the base of this analysis, it can be decided what
characteristics are going to be optimal for description of the given process.
As we have already mentioned above, when
thinking about optimising the company processes, it is also necessary to take
into account current national and company
culture and required changes of the
company culture. The more we would like to respect current national and
company culture, the looser we can define company processes. However, as soon
as we desire to change the company culture then, the stronger the change, the
more rigid description of processes is appropriate. This more rigid approach
gives exact advice how to solve the problems. The more rigid description should
be binding until new values and principles of new company culture become
absolutely familiar to all employees.
We have to take into account the considerations
about national and company culture also when we are choosing the right software
support for our processes. The majority of standard software has already
explicitly or implicitly encoded culture of country of their origin (e.g. form
of communication, meaning of certain symbols, priorities of certain values,
context data), as described by Kay and Little (Kaye&Little, 2000) or Moores and Gregory (Moores&Gregory, 2000). Now it is a question whether this culture is
understandable or acceptable everywhere and by all users.
Managing the information system of a company
and systems integration must more precisely and flexibly respond to changes in
social and economic environment. This leads to an increasing number of factors
(dimensions) that have to be taken into account when developing IS/IT (time,
data, processes, user interface, software, hardware, organisation, legislation,
finance, qualification, etc.). Other important factors that must be considered
when thinking about implementing and running information system are company and
national culture and knowledge. These factors strongly influence the optimal
features of company processes.
In order to be able to manage IS/IT even when
the number of considered dimensions grows, it is helpful to create concepts
that make the basic principles of management clear. This article has tried to
contribute to our understanding in this direction. It describes basic concept
of systems integration. This concept points out the necessity and possibility
of integration of company processes from the following three points of view:
·
Integration of processes on different levels of
management (beginning at strategic
level and finishing with operational processes),
·
Integration of processes of different areas of
management (business logistics, organisation, finance, IS/IT),
·
Integration of processes with business and
individual knowledge and with company culture.
The integration of processes with business and individual knowledge can
be managed through appropriately chosen level of business processes
definition. We have described 4
possible levels of process definition. The first is the briefest, the fourth,
on the other hand, defines the process into very detail. The more we are
approaching the fourth level, then:
·
the more accumulated business knowledge is
required (higher levels of knowledge, experience and creativity of employees
defining the process)
·
the lower are the requirements on qualification
and creativity of employees realising the process
·
the lower the flexibility of the process to
external changes
·
the higher the standardisation of the
process and standardisation of process
outputs
·
the higher the cost of process definition and
the lower the cost of process realisation
·
the better the estimates of time period and
costs of the process
·
the more appropriate are standardised
application packages such as ERP. On the other hand, the more we are
approaching the first level, the more appropriate are universal tools such as
table processor (application of business intelligence).
REFERENCES
Basl, J., Voříšek, J. (1999)
Socioinformatics - Trends in the Field of Business Processes Optimisation,
Proceedings of conference "BIT generative futures", Manchaster
University, November, ISDN 0905304306
Bruckner, T., Voříšek, J. (1998)
Outsourcing informačních systémů, Ekopress, Praha, ISBN 80-86119-07-6
[Information System Outsourcing]
CMM (1995) Http://rbse.jsc.nasa.ogv/proces_maturity/CMM/TR25
Davenport, T. (1996) Some
principles of Knowledge Management, Austin, Texas, http://
www.bus.utexas.edu/kman/kmprin.htm
Dohnal, J., Pour, J. (1997)
Architektury informačních systémů, Ekopress, Praha, [Information System
Architectures]
Donovan,J.J. (1994) Business
Re-engineering with Information Technology, Prentice Hall, ISBN 0-13-125907-5
Drucker, P.F. (1999) Management
Challenges for the 21st Century, Harpebusiness, 1999, ISBN 0887309984
Hesselbein, F., Cohen, P. (1999)
Leader to Leader, Drucker Foundation Books, ISBN 0787947261
Kaye, R., Little, S. (2000)
Dysfunctional Development Pathways of Information and Communication Technology
- Curtural Conflicts, Journal of Global Information Management, Vol. 8, No.1
January-March 2000, 5-13
Moores, T.T., F.H. Gregory (2000)
Curtural Problems in Applying SSM for IS Development, Journal of Global
Information Management, Vol. 8, No.1 January-March 2000, 14-19
Scheer, A.W. (1997) Zvyšování
produktivity pomocí podnikového modelování metodou ARIS, Systémová integrace,
ročník 4, číslo 4, ČSSI, Praha, prosinec, ISSN 1210-9479
Scheer, A.W. (1999) ARIS - od
podnikových procesů k aplikačním systémům, Comsoft ČR, Brno, ISSN
80-238-4719-8 [ARIS - From Business Processes to Application Systemes ]
Pour J., Voříšek J. - editors
(1999) System Integration '99, Proceedings of 7th International Conference,
VSE, Prague, 1999, ISBN 80-7079-059-8
Thurow, L.C. (1997) The Future of
Capitalism - How Today's Economic Forces Shape Tomorrow's World, Pengium, ISBN
0140263284
Voříšek, J, (1997) Strategické
řízení a systémová integrace, Management Press, Praha, [Strategic Management
and Systems Integration]
Voříšek, J. (1998) IS/IT Aspects of
A Co-operative Society, in Business Information Technology Management - Closing
the International Divide, Proceedings of BIT World Conference, Delhi, ed. P.
Benerjee, R. Hackney, G. Dhillon, R. Jain, Har-Anand Publications PVT LTD, New
Delhi
Voříšek, J. (1999) Systémová
integrace na prahu nového tisíciletí, Systémová integrácia'99, sborník
mezinárodní konference, Tatranská Lomnica, TU Žilina, 1999, 7-17, viz též http://nb.vse.cz/~vorisek/, (Systems
Integration at the Treshold of a New Milenium)