Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
JOURNAL OF SYSTEMS, MAN, AND CYBERNETICS - PART C: APPLICATION AND REVIEWS, VOL. X, NO. X, MONTH YEAR 1
Unleashing the Effectiveness of Process-orientedInformation Systems: Problem Analysis,
Critical Success Factors, ImplicationsBela Mutschler, Manfred Reichert, and Johannes Bumiller
Abstract—Process-oriented information systems (IS) aim at the computerized support of business processes. So far, contemporary IS have often failto meet this goal. To better understand this drawback, to systematically identify its rationales, and to derive critical success factors for business processsupport, we conducted three empirical studies: an exploratory case study in the automotive domain, an online survey among 79 IT professionals, andanother online survey among 70 business process management (BPM) experts. This paper summarizes the findings of these studies, puts them inrelation with each other, and uses them to show that ”process-orientation” is scarce and ”process-awareness” is needed in IS engineering.
Index Terms—Process-oriented Information Systems, Business Process Support, Critical Success Factors, Case Study, Online Survey.
✦
1 MOTIVATION
Providing effective IT support for business processes has be-come crucial for enterprises to stay competitive in their market[1]. In the automotive domain, for example, a broad spectrumof business processes, ranging from simple administrativeprocedures to very complex, knowledge-intense engineeringprocesses has to be effectively supported [2], [3]. Similarscenarios exist in many other domains like e-commerce [4],transportation [5], or healthcare [6]. In all these cases, domain-specific processes must be defined, implemented, enacted,monitored, and continuously adapted to a changing context.Thus, process life cycle support[7] and continuous processimprovement adopt a key role in contemporary and futureenterprise computing.
The process life cycle (cf. Fig. 1) starts with the (re)de-sign of a business process. Process modeling and processanalysis tools can be used during this phase. Thereafter,the business process has to be implemented resulting in aprocess-oriented IS. As a typical example consider aproductdata management(PDM) system which offers a broad rangeof business functions to deploy models and documentationof the managed product(s) to involved user groups (e.g.,engineers, managers, suppliers). Following the implementationand deployment phase, multiple instances of the implementedbusiness process can be created and executed during the enact-ment phase. Finally, process enactment logs can be analyzedand mined in the diagnosis phase to identify potentials forprocess optimizations.
This paper focuses on the implementation phase, i.e., thedevelopment and maintenance of process-oriented IS. Weinvestigate why contemporary IS often fail to provideeffective
• B. Mutschler and J. Bumiller are with the Department ProcessDesign, DaimlerChrysler Group Research, Ulm, Germany (email:{bela.mutschler;johannes.bumiller}@daimlerchrysler.com).
• M. Reichert is with the Information Systems Group, University of Twente,The Netherlands (email: [email protected]).
Manuscript received November 28, 2006; revised March 14, 2007.
Process
Design
Process
Implementation
Process
Enactment
Process
Diagnosis
Fig. 1. Process Life Cycle [7].
business process support. We will consider business processsupport as beingeffective, if the following two goals areachieved: (1) the cost-effective implementation or customiza-tion of processes, and (2) the availability of a technicalinfrastructure supporting all phases of the process life cycle.Based on this, our empirical work has been guided by thefollowing two research questions:
• Research Question 1: What are the major problemsleading to ineffective process support by IS?
• Research Question 2: What are critical success factorsenabling effective process support by IS?
In order to cope with these research questions, we conductedthree empirical studies (cf. Fig. 2): one exploratory case studyand two online surveys. Thecase studywas accomplished inthe automotive domain. Its goal was to identify major problemareas derogating the development and operational use of IS inpractice. Following this case study, we accomplished anonlinesurveyamong IT professionals in order to analyze and studyselected findings of our case study in more detail. Finally, weconducted anotheronline surveyamong BPM experts to derivecritical success factors for process-oriented IS. While thefirsttwo studies addressResearch Question 1, the second surveyis associated withResearch Question 2(cf. Fig. 2).
This work has been conducted in the EcoPOST project[8], [9], [10]. This project deals with the development ofa comprehensive methodology, which enables IS engineers
Study 1 Deriving the Baseline
Step 1: Interviews with Domain
Experts and IS Users
Step 2: Analysis of Process &
Software Documentation
Step 3: Derivation of Shortcomings
and Problem Areas
Engineers, Draw Chek-
kers, Documentarists, etc.
System Handbooks, Pro-
cess Instructions, etc.
>120 Critical Items,
5 Problem Areas
Step 5: Cross-organizational
Survey involving BPM Experts
Step 4: Cross-organizational
Survey involving IT Professionals
79 Participants from more
than 65 Enterprises
Exploratory
Case Study
70 Participants from more
than 55 Organizations
First
Survey
Second
Survey
1
2
3
Study 2 Detailed Analysis of Selected Issues
Study 3 Critical Success Factor Analysis
Research
Question 1
Research
Question 2
Consecutive Accomplishment
Steps: Information Sources:
Case Study Results:
Information Source:Online Survey:
Information Source:Online Survey:
Fig. 2. Three Empirical Studies.
to model and evaluate the complex economics of process-oriented IS. In particular, we focus on the interplay betweentechnological, organizational, and project-specific evaluationfactors and resulting effects.
The remainder of this paper is organized as follows. Section2 summarizes findings of the case study and details thembased on results of the first survey. Section 3 summarizesresults of the second survey. Section 4 discusses our overallfindings and explains why ”process-orientation” is scarce and”process-awareness” is needed in IS engineering. Section 5shows how we utilized our results in the EcoPOST project.Section 6 discusses related work, and Section 7 concludes witha summary and an outlook.
2 PROBLEM INVESTIGATION
This section summarizes and discusses major findings of ourcase study and our first survey.
2.1 Deriving the Baseline: An Exploratory Case Study inthe Automotive Domain
Over a period of three months we analyze two characteristicautomotive processes (a release management process and adata retrieval process) as well as their IT support (e.g., bya PDM system with more than 5000 users). We conduct26 interviews with software developers, domain experts, andend users. The interviews are based on a predefined, semi-structured protocol comprising two parts. The first one ad-dresses the investigated process, whereas the second part dealswith specific problems of the supporting IS. Besides, weanalyze process documentation and organizational handbooks.
Altogether, we collect more than 120 shortcomings relatedto the development and operational use of the investigated IS.These included both organizational and technological aspects.
Due to lack of space, we cannot present all of these 120shortcomings in detail. Instead, we summarize major resultsalong five mainproblem areas:
• Problem Area 1: Process Evolution. According to ourcase study, many problems are related to the evolution ofbusiness processes and their variability. In the analyzeddomain, frequent process changes require the continuousadaptation of the supporting IS. However, realizing suchadaptations is a difficult task to accomplish (cf. ProblemArea 2 and Problem Area 3).
• Problem Area 2: Hard-coded Process Logic. The ana-lyzed IS exhibit a ”hard-coded” process logic, i.e., processlogic is hidden in the application code and is not sepa-rately managed, e.g., by aworkflow management system(WfMS). Each time a business process changes, deepinspections and customizations of source code modulesbecome necessary. This, in turn, results in large effortsand inefficient IS adaptations.
• Problem Area 3: Complex Software Customizing. Theanalyzed IS are realized based on standard softwarecomponents. Insufficient customization features of thesecomponents also result in an ineffective adaptation ofprocess changes. In particular, existing software compo-nents lack possibilities to customize process logic at asufficiently flexible and detailed level. This complicatesthe alignment of process-oriented IS to organization-specific requirements.
Besides these fundamental problem areas, we have iden-tified two additional problem areas. These are not directlyrelated to the development of IS, but can be linked to theanalysis of requirements prior to IS implementation:
• Problem Area 4: Inadequate Business Functions. Ourcase study reveals that provided business functions donot effectively support business processes. Many of theimplemented business functions are never used and aretherefore without any ”value”. Other business functionsprovide more functionality than actually needed. Also,business functions which are actually needed for processsupport are missing, making the automation of certainprocess activities impossible.
• Problem Area 5: Missing Process Information. Some ofthe analyzed IS log event-based execution data or statusinformation (e.g., related to the start and completion ofprocess activities). However, the structure of log datadiffers from system to system. Hence, keeping track of theprocesses or mining them generates large efforts (e.g., fornormalizing available data). In any case, missing processinformation makes it difficult to identify possible processoptimizations, process cycle times are longer than needed,and resources are not allocated in a cost-effective way.
In summary, our exploratory case study has provided initialinsight into many practical problems related to the develop-ment and maintenance of IS.
2.2 A Detailed Analysis of Selected Case Study Findings:Results from an Online Survey
In order to investigate the most-relevant results from ourdomain-specific case study in detail, we performed an ad-ditional (online) survey. Selected results of this survey aredescribed in the following.
Background Information . The survey does not only in-volve IT professionals from the automotive domain, but alsofrom other organizations and domains as well, such aspublicservices, public transportation, andsteel industry(cf. Fig. 3).79 IT professionals (equating to a response rate1 of 20.2%)from more than 65 companies from Germany, Austria, andSwitzerland have participated.
Most survey participants are IT consultants or softwareengineers. Others work in the field of IT management or ITcontrolling. Quality management, project management, andprocess design are represented as well.
0
5
10
15
20
25
A B C D E F G H I J K L
A:01 (01.27%) telecommunication
B:22 (27.85%) IT
C:22 (27.85%) IT consulting
D:04 (05.06%) automotive
E:00 (00.00%) aerospace
F:03 (03.80%) pharmaceutical/chemical
G:02 (02.53%) engineering
H:09 (11.39%) financial sector
I :01 (01.27%) energy sector
J:01 (01.27%) service sector
K:01 (01.27%) industrial research
L:11 (13.92%) other
absolute nominations*
*Two participants did not answer this question.
Fig. 3. Survey Background Information.
The questionnaire has been distributed via a Web-baseddelivery platform and comprises 29 questions. Most questionsare structured, i.e., they provide a predefined set of possibleanswers. Some questions additionally allow to denote otherthan predefined answers. Some questions also allow to denotemultiple answers (such questions are designated with ”* ” inthe subsequent Figures, e.g., in Fig. 7).
Survey Results. We first asked the survey participantswhether the current degree of process-orientation is sufficient.25.32% of the participants state that ISonly partly providea sufficient degree of process-orientation (cf. Fig. 4). 8.86%even state that current IS do not provide a sufficient degree ofprocess orientation at all. 29.11% of the participants considerthe realized process support neither as problematic nor as ad-vantageous. Only 32.92% of the participants consider availablebusiness process support as (largely or completely) sufficient.
One of the problem areas identified during our case studyconcerns process evolution (cf. Problem Area 1). Surveyresults confirm that the need to continuously adapt IS to
1. Mehta and Sivadas [11] describe that response rates for electronic surveysrange from 40% to 64%. Bachmann et. al [12] identify response rates of 19%for email and 46% for mail surveys. Falconer and Hodgett [13] note thatreasonable response rates for IS research are likely to be inthe range of10% to 35%. Thus, given the low response rates to IS and email surveys ingeneral, and the large number of 29 questions, we regard the response rate toour survey as acceptable.
0
5
10
15
20
25
30
Question: Do process-oriented IS provide a
sufficient degree of business process support?
A B C D E F
absolute nominations
A:01 (01,27%) yes
B:25 (31,65%) largely
C:23 (29,11%) indifferent
D:20 (25,32%) only partly
E:07 (08,86%) no
F:03 (03,80%) don’t know
Fig. 4. Degree of Process Support.
evolving processes constitutes a problem in practice. 43.04%of the survey respondents (cf. Fig. 5) answer the question,whether their current enterprise IS can be adopted to evolvingbusiness processes (and therefore to evolving requirements)quickly enough withno (2.53%) oronly partly(40.51%). Only2.53% answer this question withyes, 27.85% withlargely.
0
5
10
15
20
25
30
35
Question: Can process-oriented IS be adopted to
evolving business processes quickly enough?
A B C D
A:02 (02,53%) yes
B:22 (27,85%) largely
C:18 (22,78%) indifferent
D:32 (40,51%) only partly
E:02 (02,53%) no
F:03 (03,80%) don’t know
E F
absolute nominations
Fig. 5. Information System Adaptations.
More than 90% of the participants agree that businessprocesses changevery often, often or sometimesin theirorganization (cf. Fig. 6A).
0
10
20
30
40
50
60
Question: Will the frequency of business process change
increase in future when compared to today?
A:54 (68,35%) increase
B:18 (22,78%) indifferent
C:05 (06,33%) decrease
D:02 (02,53%) don’t know
ab
so
lute
no
min
ati
on
s
A B C D
0
5
10
15
20
25
30
35
40
Question: How often do business processes
change in your organisation?
A B C D E F
A:02 (02,53%) very often
B:36 (45,57%) often
C:36 (45,57%) sometimes
D:04 (05,06%) rarely
E:00 (00,00%) never
F:01 (01,27%) don’t know
ab
so
lute
no
min
ati
on
s
A
B
Fig. 6. Process Changes.
Additionally, 68.35% believe that the frequency of businessprocess changes will increase in future (cf. Fig. 6B).
We also analyzed drivers for process evolution. Participantsstate that the need forprocess optimization(65.82%) is therebythe most important driver (cf. Fig. 7). Others areorganiza-tional engineering(49.37%),compliance issues(46.84%), andmarket dynamics(49.37%).
Question: What are factors leading to business processes evolution?
0
10
20
30
40
50
60
A B C D E F G H I J K L M N O P
absolute nominations
A:39
B:37
C:52
D:39
E:26
F:22
G:20
H:08
I :03
J:15
K:08
L:18
M:04
N:26
O:00
P:02
(49,37%)
(46,84%)
(65,82%)
(49,37%)
(32,91%)
(27,85%)
(25,32%)
(10,13%)
(03,08%)
(18,99%)
(10,13%)
(22,78%)
(05,06%)
(32,91%)
(00,00%)
(02,53%)
organizational engineering
laws and policies (compliance)
process optimizations
market dynamics
management order
change of enterprise goals
new software technologies
new hardware technologies
compatibility with suppliers
compatibility with customers
norms and standards
high process complexity
low user acceptance
quality program
don’t know
others
Fig. 7. Drivers of Evolution*.
Besides, we also investigate the problem of inadequatebusiness function support in more detail (cf. Problem Area4). 45.57% of the respondents share the opinion that businessprocess requirements (specifying which business functions areto be implemented) must be considered when developing anIS (cf. Fig. 8A).
0
10
20
30
40
50
Question: The requirements and needs of the business
processes are currently considered when developing
respectively customizing process-oriented IS?
A B C D
A:07 (08,86%) yes
B:42 (53,16%) largely
C:15 (18,99%) indifferent
D:13 (16,46%) only partly
E:00 (00,00%) no
F:02 (02,53%) don’t know
E F
ab
so
lute
no
min
ati
on
s
0
510
1520
25
3035
40
Question: The requirements and needs of the business
processes should be considered when developing
respectively customizing process-oriented IS?
A B C D
A:36 (45,57%) yes
B:33 (41,77%) if possible
C:07 (08,86%) indifferent
D:01 (01,27%) only partly
E:01 (01,27%) no
F:01 (01,27%) don’t know
E F
ab
so
lute
no
min
ati
on
s A
B
Fig. 8. Considering Process Requirements.
41.77% state that respective requirements should be consid-ered if possible. Therewith, 87.34% of the participants expectbusiness process requirements to be considered when imple-menting an IS. However, and this is important, only 62.02%of the participants acknowledge that respective requirementsare indeed (yes and largely) considered when developing IS(cf. Fig. 8B).
2.3 Discussion
The results of both our case study and our online surveyshow that current IS are unable to provide business processsupport as needed in practice. In our case study, we haveidentified five major reasons for this drawback: (i) continuousevolution of business processes, (ii) hard-coded process logicof the supporting IS, (iii) complex software customization,(iv) inadequate business functions, and (v) missing processinformation.
Our survey confirms these problems. Moreover, it providesfurther insights, e.g., into the drivers of process evolution orthe compliance of IS with process requirements. Picking upResearch Question 1, our results provide insights into criticalissues aggravating the introduction and operational use ofIS.
3 CRITICAL SUCCESS FACTOR ANALYSIS
In order to derivecritical success factors(CSF) for achievingeffective business process support (cf.Research Question 2),we have conducted a second online survey. As we wantto identify success factors from a business process supportperspective, only BPM experts have been involved in thissurvey. This section gives background information about thissurvey, presents major results, and discusses them.
3.1 Background Information
We performed our second survey over a period of two monthsin 2006. Like the first one, it was distributed via a Web-basedquestionnaire. The number of 70 participants corresponds to aresponse rate of 26.21%. Fig. 9 gives background informationabout survey participants.
The questionnaire is based on a profound literature study oncritical success factors(CSF) for IS implementation in general(cf. Section 6) as well as on the results of the case study andour first online survey.
3.2 Survey Results
Basic to the survey2 is the distinction between organization,project-, and technology-specific CSFs.
Organization-specific Critical Success Factors. Organi-zation-specific CSFs deal with attributes of an organizationthat bias the development of process-oriented IS. As anexample considerdomain knowledgeand its positive impacton the redesign of business processes. Another example isprovided byorganizational process maturity. Process maturitycan be assessed by dedicated maturity models that describecharacteristics of effective process organizations. Examples arethe capability maturity model integration(CMMI) [15] or thesoftware process improvement and capability determination(SPICE) model [16]. If the process maturity of an organizationis low, it will be more difficult to implement (optimized)business processes.
In our survey,end user participation(47.14%) andaccess torequired information(42.86%) are those organization-specificCSFs that aggregate most nominations as ”essential factor”
2. We have described the complete findings of this survey in [14].
0
5
10
15
20
25
30
35
40
A B C D E FQuestion: How would you rate your own knowledge regarding BPM?
absolute nominations
A:16
B:36
C:12
D:05
E:00
F:01
(22.86%)
(51.43%)
(17.14%)
(07.14%)
(00.00%)
(01.43%)
expert knowledge
good knowledge
some knowledge
little knowledge
no knowledge
don’t know
0
5
10
15
20
25
30
A B C D EQuestion: How long are you working in the field of BPM?
absolute nominations
A:25
B:20
C:16
D:07
E:02
(35.71%)
(28.57%)
(22.86%)
(10.00%)
(02.86%)
>5 years
>3 years
>1 year
<1 year
don’t know
05
10
1520253035
404550
A B C D EQuestion: What is your background?
absolute nominations
A:46
B:04
C:12
D:00
E:08
(65.71%)
(05.71%)
(17.14%)
(00.00%)
(11.43%)
university
industrial research
industrial
don’t know
other
A
B
C
Fig. 9. Survey Background Information.
(cf. Fig. 10). Factors that are considered as ”very important”are as follows:reorganization of information, availability ofprocess documentation, ability to redesign business processes,and the ability of an organization to adapt its IT governance.In order to better understand the relevance of the analyzedevaluation factors, Fig. 11 shows the mean for each CSF.
0
5
10
15
20
25
30
35
Q 1 Q 2 Q 3 Q 4 Q 5 Q 6 Q 7 Q 8 Q 9 Q 10
Question: Evaluate the following ORGANIZATION-specific criticalsuccess factors regarding their importance for the use of BPM technology.
absolu
te n
om
inations
Q 1
Q 2
Q 3
Q 4
Q 5
Q 6
Q 7
Q 8
Q 9
Q 10
A
(22.86%)
(34.29%)
(18.57%)
(32.86%)
(47.14%)
(42.86%)
(12.86%)
(14.29%)
(10.00%)
(17.14%)
Experiences in using BPM technology
Ability to redesign business processes
Reorganization of information
Domain knowledge
End user participation
Access to required information
Organizational process maturity
Availability of process documentation
Mature technology infrastructure
Ability to adapt the IT governance
E
(08.57%)
(07.14%)
(07.14%)
(11.43%)
(05.71%)
(07.14%)
(07.14%)
(05.71%)
(08.57%)
(08.57%)
D
(01.43%)
(00.00%)
(01.43%)
(05.71%)
(01.43%)
(00.00%)
(10.00%)
(07.14%)
(12.86%)
(10.00%)
C
(41.43%)
(20.00%)
(24.29%)
(12.86%)
(15.71%)
(17.14%)
(32.86%)
(44.29%)
(44.29%)
(35.71%)
B
(25.71%)
(38.57%)
(48.57%)
(37.14%)
(30.00%)
(32.86%)
(37.14%)
(28.57%)
(24.29%)
(28.57%)
A essential B very important C important D unimportant E don’t know
Fig. 10. Organizational CSFs.
Project-specific Critical Success Factors. Project-specific
End user participation
10 3
unim
port
ant
import
ant
2
very
im
port
ant
Experiences in using BPM technology
Ability to redesign/reengineer business processes
Reorganization of information
Domain knowledge
Access to required information
Organizational process maturity
essential
Availability of process documentation
Mature technology infrastructure
Ability to adapt the IT governance
Organization-specific
Critical Success Factors
(shows the mean
for each factor)
Mean V
alu
es
Fig. 11. Organizational CSFs (Means).
CSFs deal with project-driven attributes which influence thedevelopment of process-oriented IS. As examples considerfactors such asknowledge about existing processesor accessto required skills.
Question: Evaluate the following PROJECT-specific critical successfactors regarding their importance for the use of BPM technology.
0
5
10
15
20
25
30
35
40
45
50
Q 1 Q 2 Q 3 Q 4 Q 5 Q 6 Q 7 Q 8 Q 9 Q 10 Q 11 Q 12 Q 13
absolute nominations
A essential B very important C important D unimportant E don’t know
Q 1
Q 2
Q 3
Q 4
Q 5
Q 6
Q 7
Q 8
Q 9
Q 10
Q 11
Q 12
Q 13
A
(35.71%)
(32.86%)
(25.71%)
(11.43%)
(04.29%)
(02.86%)
(20.00%)
(45.71%)
(67.14%)
(15.71%)
(25.71%)
(22.86%)
(38.57%)
Overview of existing processes
Knowledge about existing processes
Information about existing processes
Evolutionary process redesign
Revolutionary process redesign
Degree of job redesign
End user fears
Communication with end users
Management commitment
Use of process modeling tools
Adequate planning
Access to required skills
Motivation for the project
E
(11.43%)
(10.00%)
(10.00%)
(14.29%)
(17.14%)
(18.57%)
(18.57%)
(10.00%)
(10.00%)
(08.57%)
(08.57%)
(10.00%)
(08.57%)
D
(01.43%)
(01.43%)
(00.00%)
(08.57%)
(15.71%)
(12.86%)
(02.86%)
(01.43%)
(01.43%)
(02.86%)
(02.86%)
(02.86%)
(01.43%)
C
(20.00%)
(18.57%)
(32.86%)
(44.29%)
(44.29%)
(45.71%)
(17.14%)
(12.86%)
(05.71%)
(40.00%)
(28.57%)
(27.14%)
(18.57%)
B
(31.43%)
(37.14%)
(31.43%)
(21.43%)
(18.57%)
(20.00%)
(41.43%)
(30.00%)
(15.71%)
(32.86%)
(34.29%)
(37.14%)
(32.86%)
Fig. 12. Project-specific CSFs.
According to our survey (cf. Fig. 12),management commit-ment (67.14%) andcommunication with end users(45.71%)aggregate most nominations as ”essential” project-specificfactors. Several other CSFs are considered as ”very important”,including degree of job redesign, information about existingprocesses, andproject motivation. Fig. 13 shows the mean foreach CSF.
Technology-specific Critical Success Factors. Technolo-gy-specific CSFs deal with the technical infrastructure fordeveloping and maintaining IS. As an example consider thedegree of flexibilityprovided by an IS (e.g., regarding its abilityto allow for dynamic process changes at run-time; cf. ProblemArea 1). Another example is the use ofstandardsfor processspecification (e.g., BPMN, WS-BPEL, BPML).
10 32
unimportant
important
very important
essential
Revolutionary process redesign
Overview of existing processes
Knowledge about existing processes
Information about existing processes
Evolutionary process redesign
Degree of job redesign
End user fears
Communication with end users
Management commitment
Use of process modeling tools
Adequate planning
Access to required skills
Motivation for the project
Project-specific Critical
Success Factors
(shows the mean
for each factor)
Mean Values
Fig. 13. Project-specific CSFs (Means).
Finally, consider the availability ofdevelopment tools.Recently,business process intelligence(BPI) tools [17] areoften discusses in this context. BPI tools analyze event-basedprocess execution data (e.g., start and completion times ofprocess activities, resources consumed by a process activity,process cycle times).
Question: Evaluate the following TECHNOLOGY-specific critical successfactors regarding their importance for the use of BPM technology.
0
5
10
15
20
25
30
Q 1 Q 2 Q 3 Q 4 Q 5 Q 6 Q 7 Q 8 Q 9 Q 10 Q 11 Q 12
absolute nominations
A essential B very important C important D unimportant E don’t know
Q 1
Q 2
Q 3
Q 4
Q 5
Q 6
Q 7
Q 8
Q 9
Q 10
Q 11
Q 12
A
(12.86%)
(10.00%)
(12.86%)
(12.86%)
(07.14%)
(14.29%)
(04.29%)
(17.14%)
(05.71%)
(05.71%)
(20.00%)
(04.29%)
Technical maturity of the BPM platform
Experiences using the BPM platform
Available support for the BPM platform
Supported degree of process flexibility
Availability of developing tools
Support of standards and norms
Low license costs
Good documentation
Regular product updates
Model-driven application development
Usability of the BPM platform
Powerful application programming interface
E
(12.86%)
(11.43%)
(11.43%)
(11.43%)
(15.71%)
(11.43%)
(12.86%)
(12.86%)
(12.86%)
(14.29%)
(11.43%)
(14.29%)
D
(04.29%)
(11.43%)
(07.14%)
(08.57%)
(08.57%)
(14.29%)
(34.29%)
(04.29%)
(30.00%)
(15.71%)
(04.29%)
(17.14%)
C
(38.57%)
(35.71%)
(32.86%)
(27.14%)
(34.29%)
(32.86%)
(35.71%)
(27.14%)
(38.57%)
(37.14%)
(27.14%)
(32.86%)
B
(31.43%)
(31.43%)
(35.71%)
(40.00%)
(34.29%)
(27.14%)
(12.86%)
(38.57%)
(12.86%)
(27.14%)
(37.14%)
(31.43%)
Fig. 14. Technological CSFs.
Interestingly, none of the listed technology-specific CSFsis considered as essential by the majority of the surveyrespondents (cf. Fig. 14).Good documentation(17.14%) andusability (20%) get most nominations as ”essential factors”.Moreover, these two factors are considered as ”very important”by many survey participants (38.57% and 37.14%).
Several CSFs are considered as ”very important” or ”im-portant”. These CSFs include:available vendor support for aBPM systemandavailability of suitable development tools.
Note that the number of survey respondents giving noanswer (i.e., denoting ”don’t know”) does only slightly varyalong the analysis of technology-specific CSFs (cf. Fig. 14).This indicates that some survey participants might not havebeen able to interpret the listed CSFs. Fig. 15 shows the meanfor each CSF.
Availability of developing tools
Technical maturity of the BPM platform
Experiences using the BPM platform
Available support for the BPM platform
Supported degree of process flexibility
Support of standards and norms
Low license costs
Good documentation
Regular product updates
Model-driven application development
Usability of the BPM platform
Powerful application programming interface
10 32
unimportant
important
very important
essential
Technology-specific
Critical Success Factors
(shows the mean
for each factor)
Mean Values
Fig. 15. Technological CSFs (Means).
3.3 Discussion
Based on the results of the second survey, we answerResearchQuestion 2(cf. Section 1). Thereby, it seems hardly possibleto consider all technology-specific CSFs when introducing anIS. In any case, these CSFs represent a validated baselinefor enterprises that want to increase the effectiveness of theirprocess-oriented IS.
Basic to our survey is the distinction between organization-, project-, and technology-specific CSFs. It is possible tocategorize CSFs in another way (e.g., using more technicalcategories). We consider the chosen categorization as the mostuseful one. In particular, it provides an intuitive baseline thatcan be easily understood by IT professionals (including oursurvey participants).
To ensure internal validity of our survey results we havedeveloped the questionnaire based on the results of the pre-ceding case study and online survey as well as on a literaturestudy on CSFs for IS implementation (cf. Section 6).
To increase external validity, only BPM experts may par-ticipate, i.e., we exclude answers of respondents having no”process background” since the ”assumed answers” of theseparticipants would falsify results. In this context, it is alsoimportant to mention that the survey questionnaire has allowedto denote other than the predefined answers. However, thisopportunity was not used.
4 IMPLICATIONS
Based on results of ourfirst survey, we take a closer look at thedissemination of process-oriented software technologiesandexplain why ”process-orientation” is scarce and why ”process-awareness” is needed in practice.
4.1 Process-oriented Software Technologies
In recent years, many process support paradigms (e.g., work-flow management, service orchestration and service choreogra-phy, case handling), process specification standards (e.g., WS-BPEL, BPML, BPMN), and BPM tools (e.g., ARIS Toolset,Tibco Staffware, FLOWer) have emerged [18]. Their goal isto realize effective IT support for business processes. In ourfirst survey, we have investigated the dissemination of theseapproaches in more detail.
63.29% of the respondents confirm the use ofWfMS (cf.Fig. 16). 36.71% even deploy more comprehensiveBPMsystems. 53.16% of the respondents confirm the (exploratory)use ofweb services. 32.91% already set up completeservice-oriented architectures. Enterprise application integrationplat-forms (e.g., to enable process integration) are applied by39.24% of the respondents.
These results do not allow for any conclusion regardingthe extent to which enterprises use a respective technology.Nevertheless, the increasing importance and use of process-oriented software technologies is indicated.
0
10
20
30
40
50
60
A B C D E F G HQuestion: Which software technologies are
used in your organization today?
I J K L M
absolute nominations
A:50
B:29
C:31
D:26
E:19
F:25
G:53
H:36
I :24
J:42
K:27
L:05
M:03
(63,29%)
(36,71%)
(39,24%)
(32,91%)
(24,05%)
(31,65%)
(54,43%)
(45,57%)
(30,38%)
(53,16%)
(34,18%)
(06,33%)
(03,80%)
Workflow management systems
BPM systems
Enterprise application integration
Service-oriented architectures
Rule-based systems
Process portals
Knowledge and enterprise portals
Data warehousing technologies
Business intelligence
Web services
Collaboration tools
Don’t know
Others
Fig. 16. Technologies used Today*.
In order to investigate the sustainability of this trend, wehave asked survey participants about which process-orientedsoftware technologies they will use in future (cf. Fig. 17).Themost frequent answers areWfMS(39.24%),services-orientedarchitectures(36.71%),web services(32.91%), andprocessportals (30.38%). This indicates a continuing disseminationof process-oriented software technologies.
4.2 Process-aware Information Systems
Our empirical studies (cf. Fig. 18) indicate that providingeffective business process support by IS is a difficult task
Question: Which software technologies are
planned to be used in your organization in the future?
0
5
10
15
20
25
30
35
A B C D E F G H I J K L M
absolute nominations
A:31
B:22
C:17
D:29
E:12
F:24
G:21
H:21
I :23
J:26
K:17
L:16
M:04
(39,24%)
(27,85%)
(21,52%)
(36,71%)
(15,19%)
(30,38%)
(26,58%)
(26,58%)
(29,11%)
(32,91%)
(21,52%)
(20,25%)
(06,06%)
Workflow management systems
BPM systems
Enterprise application integration
Service-oriented architectures
Rule-based systems
Process portals
Knowledge and enterprise portals
Data warehousing technologies
Business intelligence
Web services
Collaboration tools
Don’t know
Others
Fig. 17. Technologies used Tomorrow*.
to accomplish. The currently realized degree of ”process-orientation” in IS is by far not satisfactory. By contrast,enterprises more and more crave for approaches that enablethem to improve their business process performance. Reflect-ing the aforementioned results, we conclude that conventionalprocess-oriented IS are scarce. What we need instead, areprocess-aware IS(PAIS), i.e., IS that support all phases allof the process life cycle.
PAIS can be implemented in two ways [18]: (1) by de-veloping an organization-specific process support system,or(2) by configuring a generic process support system. In theformer case, the PAIS is build ”from scratch” and incorporatesorganization-specific information about the structure andpro-cesses to be supported. As an example consider anenterpriseresource planning(ERP) system. In the latter case, the PAISdoes not contain any information about the structure and theprocesses of a particular organization. Instead, an organizationneeds to configure the PAIS by specifying processes, organi-zational entities, and business objects. As an example considerthe configuration of a WfMS.
In any case, PAIS strictly separate process logic (comprisingthe activities to be executed) from application code [19], i.e.,PAIS are driven by process models rather than program code.PAIS are realized based on powerful process engines whichorchestrate processes at run-time [20]. These process enginesalso provide extensive libraries of process-oriented functions atbuild-time, e.g., for accomplishing automatic process analysis(based on BPI tools). Empirical studies [21] confirm thatPAIS enable the fast and cost-effective implementation andcustomization of new and of existing processes (cf. ProblemAreas 2 + 3).
Realizing PAIS also implies a significant shift in the fieldof IS engineering. Traditional IS engineering methods andparadigms (e.g., procedural programming) have to be supple-mented with engineering principles particularly enhancing theoperational support of business processes. This is crucialto tieup to those requirements that have been neglected by current
Organization-specific CSFs Technology-specific CSFs
Process-oriented
Information Systems
Problem
Areas (PA)Process-aware
Information Systems
Critical
Success
Factors
enable the
introduction of
Project-specific CSFs
Process
Evolution
Hard-coded
Process Logic
Complex Soft-
ware Customizing
Inadequate
Business Functions
Missing Pro-
cess Information
Case Study
Survey 2
Research Question 1: What are the major problems leading to ineffective process support by IS?
Research Question 2: What are CSFs enabling effective process support by IS?
PA1 PA2 PA3 PA4 PA5
Survey 1
Subject of
Analysis
promise to
overcome
Fig. 18. Three Empirical Studies: The Big Picture.
process-oriented IS so far. However, such a shift is difficultto accomplish as IS projects often use software technologies– at least today – that do not support the needed degree ofprocess-orientation.
5 UTILIZING OUR RESULTS IN ECOPOST
In the EcoPOST project, we are developing a framework formodeling and investigating the complex interplay between thenumerous technological, organizational and project-driven costand impact factors which arise in the context of process-oriented IS and PAIS (and which do only partly exist inprojects dealing with data- or function-centered IS).
Costs for business process redesign, for example, may beinfluenced by an intangible impact factor ”Willingness of StaffMembers to support Redesign Activities”. If staff membersdo not contribute to a redesign project by providing neededinformation (e.g., about process bottlenecks), any redesigneffort will be ineffective and will increase costs. If staffwillingness is additionally varying during the redesign activity(e.g., due to a changing communication policy), businessprocess redesign costs will be subject to more complex effects.
5.1 Motivating Example
In order to make the dependencies and the interplay betweentechnological, organizational and project-driven factors ex-plicit in the EcoPOST project, we have developed the notionof evaluation models[10]. These models are formulated usingthe System Dynamicsnotation [22]. In addition, they can besimulated in order to unfold the dynamic behavior describedby them. In the following, we abstract from formal andtechnical details regarding our evaluation models. Instead, weshow how we validate our evaluation models based on thepresent survey results along an example.
Example. Fig. 19 shows a model which describes theinfluence of an impact factor ”End User Fears” on a cost factor”Costs for Business Process Redesign”. More specifically, thismodel reflects the assumption that the introduction of a PAISmay cause end user fears, e.g., due to job redesign and dueto changed social clues. Such fears can lead to emotionalresistance of end users. This, in turn, can make it difficult toget needed support from end users, e.g., during an interview-based process analysis.
Illustrating Example: Analyzing the Role of „End User Fears“
LOOP
Degree of
Job Redesign
Change of
Social Clue and
Interactions
End User
Fears
Communication
Ability to redesign
Business
ProcessesAbility to
acquire Process
Knowledge
Emotional
Resistance of
End Users
+
-
+
+
-
+
Costs for Business
Process Redesign
+
+
Fig. 19. Modeling the Impact of User Fears.
Basic to this evaluation model is a cyclic structure (orfeedback loop) connecting the four factors ”End User Fears”,”Emotional Resistance of End Users”, ”Ability to acquireProcess Knowledge”, and ”Ability to redesign Business Pro-cesses”. Their arrangement illustrates the following coherence:increasing end user fears lead to increasing emotional resis-tance of end users. This dependency is represented by a ”pos-itive link” from ”End User Fears” to ”Emotional Resistanceof End Users” (cf. Fig. 19).
In our evaluation models, apositive linkbetween variables x
and y (with y the dependent variable) indicates that y tends inthe same direction if a change occurs in x. A negative link, bycontrast, would denote that y tends in the opposite direction.
Returning to our example, an increasing emotional resis-tance of end users may result in a decreasing ability to ac-quire process knowledge. Reason is that increasing emotionalresistance makes profound process analysis, e.g., based oninterviews with process participants, a difficult task to accom-plish. This causal dependency is represented by a negativelink pointing from ”Emotional Resistance of End Users” to”Ability to acquire Process Knowledge”.
The inability to acquire process knowledge, in turn, mayresult in a decreasing ability to redesign business processes (asneeded information is missing). Finally, an increasing abilityto redesign business processes can even enforce end userfears. Reason is that end users often consider business processredesign activities as a potential threat for their own job.
Note that the variable ”End User Fears” is not only influ-enced by the ”Ability to redesign Business Processes”, butby other variables as well, e.g., the expected ”Degree of JobRedesign” or the ”Change of Social Clue and Interactions”.Moreover, ”Communication” can decrease end user fears, e.g.,by informing end users about the goals associated with theintroduction of a PAIS. This is described by a negative linkpointing from ”Communication” to ”End User Fears”.
Model Validation . Such evaluation models (cf. Fig. 19) areof significant value for PAIS engineers. They can be used, forexample, as a starting point for building more complex eval-uation models enabling the performance of cost simulations[23], or as a means for clarifying causal dependencies in PAISengineering projects.
However, the expressiveness of such evaluation modelsdepends on the availability, plausibility and resilience of datasupporting the modeled causal dependencies. One approach toderive such data is the accomplishment of both empirical andexperimental research activities (e.g., software experiments,online surveys, case studies).
Taking the results of our second online survey (cf. Section3), for example, we can validate some of the causal depen-dencies from the evaluation model in Fig. 19.
5.2 Empirical Model Validation
In the second online survey we analyzed, for example, thecausal dependency between ”End User Fears” and ”Emotionalresistance of End Users” as well as the dependency between”Communication” and ”End User Fears”.
For this purpose, we use a four-step sequence of questionsto derive data. First, we ask for therelevanceof a given factor(Question 1). Second, we want to know whether there exists acausal dependencybetween this factor and another one (Ques-tion 2). Only those survey participants – and this is importantfor understanding related results – who answer this secondquestion with ”yes” are directed to two additional questions.These additional questions deal with the further specificationof the previously confirmed dependency. Question 3 dealswith the semantic specificationof the dependency, whereasQuestion 4 addresses thestrengthof the impact.
0
5
10
15
20
25
A B C D E
Question 4: How strong is the specified impact of this relationship?
absolute nominations
0
5
10
15
20
25
30
35
40
45
A B C D E FQuestion 3: What is the direction of this relationship?
absolute nominations
0
10
20
30
40
50
60
A B CQuestion 2: Does there exists a relationship between end user
fears and emotional resistance against BPM technology?
absolute nominations
0
5
10
15
20
25
30
35
A B C D EQuestion 1: How critical are end user fears for the success
of a BPM project / for introducing BPM technology?
absolute nominations
A
B
C
D
A:20
B:32
C:06
D:04
E:08
(28.57%)
(45.71%)
(08.57%)
(05.71%)
(11.43%)
very critical
critical
negligible
not critical
don’t know
A:49
B:06
C:15
(70.00%)
(08.57%)
(21.43%)
yes
no
don’t know
A:21B:23C:03D:00E:02
(42.86%)(46.94%)(06.12%)(00.00%)(04.08%)
very strongstrongweakvery weakdon’t know
A:41B:02C:03D:00E:02F:01
(83.67%)(04.08%)(06.12%)(00:00%)(04.08%)(02.04%)
increasing UF increasing ERincreasing UF decreasing ERincreasing ER increasing UFincreasing ER decreasing UFdon't knowthere is another relationship
UF = End User Fears
ER = Emotional Resistance
Fig. 20. The Impact of End User Fears.
Dealing with End User Fears. Consider Fig. 20. A majorityof 74.28% of the participants consider end user fears as ”verycritical” (28.57%) or ”critical” (45.71%) for the overall successof a BPM project (cf. Fig. 20A). More specifically, 70% ofthe respondents confirm that there is a relationship betweenend user fears and the emotional resistance of end usersagainst BPM technology (cf. Fig. 20B). This confirms the linkbetween ”End User Fears” and ”Emotional Resistance of EndUsers” in Fig. 19.
What is still not clear at this point is the direction of thecausal dependency – there are several possibilities in thisrespect (cf. Fig. 19). However, the semantics of the link canbe clarified based on the next question.
83.67% share the opinion that increasing end user fearsresult in increasing emotional resistance (cf. Fig. 20C). Thisconfirms that ”End User Fears” and ”Emotional Resistance of
End Users” need to be connected with a positive link pointingfrom the former variable to the latter one.
Finally, 89.8% of the respondents state (cf. Fig. 20D)that the impact of end user fears on emotional resistanceis either ”very strong” (42.86%) or ”strong” (46.94%). Thisinformation helps us to quantify the causal dependency whenwe simulate the evaluation model.
Dealing with Communication. In the evaluation modelfrom Fig. 19, we assume that increasing communication resultsin decreasing end user fears. 92.86% of the survey participantsconsider communication between the stakeholders of a BPMproject as ”essential” (47.14%), ”very important” (35.71%)or ”important” (10%) for its success. Furthermore, 78.57%of the respondents confirm that there is a causal dependencybetween communication and end user fears. Out of these,74.55% are the opinion that an increasing communicationresults in decreasing end user fears. This confirms that ”Com-munication” and ”End User Fears” can be connected with anegative link pointing from the former variable to the latterone. Finally, 85.45% of the respondents consider the impactof communication on end user fears as either ”very strong”(29.09%) or ”strong” (56.36%). This information helps uswhen specifying a corresponding EcoPOST simulation modelfor this evaluation model.
These results exemplify how we utilize survey results forvalidating (or deriving) EcoPOST evaluation models.
6 RELATED WORK
There exist several studies dealing with CSFs for imple-menting IS, mainly in the context of ERP systems. Yusufaet. al [24], for example, investigate the introduction of anERP system in a large manufacturing organization. This workfocuses on technological and cultural CSFs and comparesexpected benefits of the realized ERP system with its actualones. Vogt [25] analyzes failed ERP projects and identifies– from a software engineering viewpoint – factors whichhelp to avoid such failures. A similar study is described byVoordijk [26], who investigates ERP implementations in largeconstruction firms. Mandal et. al [27] describe experiencesgathered during during the planning and implementation stagesof an ERP implementation in a water corporation. Danevaand Wieringa use a ”success model” for reasoning aboutthe factors enabling successful ERP implementations [28].Focusing on risk factors, Sumner [29] investigates the re-engineering of processes prior to an ERP project. Discussedissues imply the recruiting and training of IT professionals, theinvolvement of consultants, and the integration of application-specific knowledge and technical expertise into the projectteam.
CSFs for workflow implementations are described by Parkes[30], [31]. Thereby, three CSFs are identified as being particu-lar important: management commitment, communication, andparticipation by end users. Note that our results confirm theimportance of these three factors.
Focusing on organizational issues, Sarmento and Machado[32] investigate the impact of workflow implementations onan organization, identify relevant contingency factors, and
describe organizational domains which are affected by aWfMS. In his qualitative study [33], Kueng also investigatesthe impact of WfMS on organizations.
Davenport [34] deals with only one CSF, namely the abilityof an organization to align its IS to business processes andbusiness strategy. Baroudi et. al [35] investigate the impact ofuser involvement on IS usage and information satisfaction.Their results demonstrate that user involvement during ISdevelopment will enhance system usage and user satisfaction.Their results also show that growing user satisfaction resultsin greater system usage.
All these studies have been considered when conducting ourempirical research. In particular, the work of Parkes [30],[31]has contributed to the design of the questionnaire for our sec-ond survey. However, none of these studies distinguishes be-tween organization-, project-, and technology-specific factors(though Parkes [30], [31] distinguishes between organizationaland technological factors).
7 SUMMARY AND OUTLOOK
Two research questions (cf. Section 1) have guided the em-pirical research presented in this paper. In this context, wehave conducted three empirical studies: an exploratory casestudy in the automotive domain and two cross-organizationalonline surveys. The case study and the first survey enableus to identify and analyze major problems related to thedevelopment, maintenance and operational use of process-oriented IS (cf.Research Question 1). The second onlinesurvey deals with CSFs for building better, i.e., more effectiveprocess-oriented IS and/or PAIS (cf.Research Question 2).
We have also discussed potential implications of our find-ings to the field of IS engineering. In particular, we have indi-cated that mere ”process-orientation” is scarce, but ”process-awareness” is needed in practice. In this context, we havealso given a short characterization of PAIS. Finally, we havedescribed the utilization of our results to validate causalde-pendencies in evaluation models of the EcoPOST framework.
Altogether, this paper provides insights into important issuesand challenges related to the introduction and use of IS inpractice. It helps both IT professionals and researchers tounderstand those factors that can improve the operationaleffectiveness of process-oriented IS.
Future work will include additional case studies, onlinesurveys, and controlled software experiments to generate ad-ditional data we can use for validating our evaluation models.Finally, we will investigate the impact of agile softwaredevelopment processes on the realization of PAIS.
REFERENCES
[1] Y. L. Antonucci, Using Workflow Technologies to ImproveOrganizational Competitiveness, Int’l. Journal of Management,14(1), pp.117-126, 1997.
[2] E. Knippel and A. Schulz,Lessons Learned from ImplementingConfig. Manag. within Electrical/Electronic Development of anAutomotive OEM, Proc. 14th INCOSE, 2004.
[3] D. Muller, J. Herbst, M. Hammori, and M. Reichert,IT Supportfor Release Management Processes in the Automotive Industry,Proc. 4th BPM, LNCS 4102, pp.368-377, 2006.
[4] J. Gordijn and H. Akkermans,Value-based Requirements En-gineering: Exploring Innovative E-Commerce Ideas, Int’l. J. ofRequirements Engineering, 8, pp.114-134, 2003.
[5] S. Bassil, Workflow Technology for Complex Socio-TechnicalSystems, PhD Thesis, University of Montreal, 2005.
[6] R. Lenz and M. Reichert,IT Support for Healthcare Processes:Premises, Challenges, Perspectives, Data and Knowledge Engi-neering, Volume 61, pp.39-58, 2007.
[7] W. M. P. van der Aalst,Business Process Management - APersonal View, BPM Journal, 10(2), pp.248-253, 2004.
[8] B. Mutschler, M. Reichert, and J. Bumiller,An Approach forEvaluating WfMS from a Value-Based Perspective, Proc. 10thIEEE EDOC, pp.477-482, 2006.
[9] B. Mutschler, M. Reichert, and J. Bumiller,Why Process-Orientation is Scarce: An Empirical Study of Process-orientedInformation Systems in the Automotive Industry, Proc. 10th IEEEEDOC, pp.433-438, 2006.
[10] B. Mutschler, M. Reichert, and S. Rinderle,Analyzing theDynamic Cost Factors of Process-aware Information Systems:A Model-based Approach, Proc. 19th CAiSE, pp.589-603, 2007.
[11] R. Mehta and E. Sivadas,Comparing Response Rates andResponse Content in Mail versus Electronic Mail Surveyed, J.of the Market Research Society, 37, pp.429-439, 1995.
[12] D. Bachmann, J. Elfrink, and G. Vazzana,E-mail and SnailMail Face Off in Rematch, Marketing Research, 11(4), pp.11-15, 2000.
[13] D. J. Falconer and R. A. Hodgett,Why Executives Don’tRespond to Your Survey, Proc. 10th ACIS, pp.279-285, 1999.
[14] B. Mutschler and M. Reichert,A Survey on Evaluation Factorsfor Business Process Management Technology, Technical ReportTR-CTIT-06-63, University of Twente, 2006.
[15] CMMI, Capability Maturity Model Integration, Software Engi-neering Institute: www.sei.cmu.edu/cmmi/, 2007.
[16] SPICE,Software Process Improvement and Capability dEtermi-nation, www.sqi.gu.edu.au/spice/, 2007.
[17] B. Mutschler, M. Reichert, and J. Bumiller,An Approachto Quantify the Costs of Business Process Intelligence, Proc.EMISA Workshop, LNI P-75, pp.152-165, 2005.
[18] M. Dumas, W. M. P. van der Aalst, and A. ter Hofstede,Process-aware Information Systems: Bridging People and Soft-ware through Process Technology, Wiley, 2005.
[19] J. Dehnert and W. van der Aalst,Bridging the Gap betweenBusiness Models and Workflow Specification, Int’l. J. of Coop.IS, 13(3), 289-332, 2004.
[20] M. Reichert, S. Rinderle, U. Kreher, and P. Dadam,AdaptiveProcess Management with ADEPT2, Proc. 21th ICDE, pp.1113-1114, 2005.
[21] N. Kleiner, Can Business Process Changes be cheaper Imple-mented with Workflow-Management-Systems?, Proc. IRMA ’04,pp.529-532, 2004.
[22] J. D. Sterman,Business Dynamics - Systems Thinking andModeling for a Complex World, McGraw-Hill, 2000.
[23] B. Mutschler and M. Reichert,Exploring the Dynamic Costsof Process-aware IS through Simulation, Proc. Workshops andDoctoral Consortium at 19th CAiSE ’07, Volume 1, EMMSADWorkshop, pp.173-182, 2007.
[24] Y. Yusufa, A. Gunasekaranb, and M. S. Abthorpe,EnterpriseInformation Systems Project Implementation: A Case Study ofERP in Rolls-Royce, Int’l. J. of Production Economics, Volume87, pp.251-266, 2004.
[25] C. Vogt, Intractable ERP: A comprehensive Analysis of failedEnterprise-Resource-Planning Projects, ACM SIGSOFT Soft-ware Engineering Notes, 27(2), pp.62-68, 2002.
[26] H. Voordijk, A. van Leuven, and A. Laan,Enterprise ResourcePlanning in a Large Construction Firm: Implementation Anal-ysis, Construction Management and Economics, 21(5), pp.511-521, 2003.
[27] P. Mandal and A. Gunasekaran,Issues in Implementing ERP:A Case Study, European Journal of Operational Research, 146,
pp.274-283, 2003.[28] M. Daneva and R. Wieringa,A Conceptual Framework for
Research in Cross-organizational ERP Cost Estimation, Proc.PROMan Workshop, 2005.
[29] M. Sumner,Risk Factors in Enterprise-wide/ERP Projects, J.of Information Technology, 15(4), pp.317-327, 2000.
[30] A. Parkes,Critical Success Factors in Workflow Implementation,Proc. 6th PACIS, pp.363-380, 2002.
[31] A. Parkes,A Case Study of Workflow Implementation SuccessFactors, Proc. 15th ACIS, 2004.
[32] A. Sarmento and A. Machado,The Adoption of WorkflowSystems: Proposal of a Model for a Methodology to Analysisthe Impact of Workflow Systems in Organizations, Proc. 2ndICEIS, pp.349-355, 2000.
[33] P. Kueng,The Effects of Workflow Systems on Organization: AQualitative Study, Business Process Management, LNCS 1806,pp.301-316, 2000.
[34] T. H. Davenport,Mission Critical: Realizing the Promise ofEnterprise Systems, 2000.
[35] J. J. Baroudi, M. H. Olson, and B. Ives,An Empirical Study ofthe Impact of User Involvement on System Usage and Informa-tion Satisfaction, Comm. ACM, 29(3), pp.232-239, 1986.
Bela Mutschler Bela Mutschler received his diplomain computer science from the University of Ulm, Ger-many, in 2003. Currently he works as a process engi-neer at DaimlerChrysler (Group Research), Germany.Bela is an expert in the fields of automotive processmanagement and software engineering, business pro-cess management (BPM), and process maturity mod-els (such as CMMI). Besides, he is an external Ph.D.student in the Information Systems Group at the Uni-versity of Twente (UT), The Netherlands, where he iscurrently finishing his Ph.D. thesis. In his dissertation,
Bela investigates causal dependencies and resulting cost effects in projectsdealing with the introduction of process-aware information systems.
Manfred Reichert Manfred Reichert holds a Ph.D. incomputer science and is currently Associate Profes-sor in the Information Systems Group at the Universityof Twente (UT), The Netherlands. At UT he is also amember of the Management Team of the Centre forTelematics and Information Technology (CTIT), whichis the largest ICT research institute in the Nether-lands. Before Manfred joined UT in January 2005,he had been assistant professor in the Institute forDatabases and Information Systems at the Universityof Ulm, Germany. There, he also finished his Ph.D
thesis on adaptive process management in July 2000 (with honours). Manfredreceived several awards for his outstanding work on the ADEPT processmanagement technology. He is an expert in the field of business process man-agement (BPM) and has conducted many BPM-related projects in applicationdomains like healthcare and automotive engineering.
Johannes Bumiller Johannes Bumiller received adiploma in computer science from the University ofKarlsruhe, Germany. Afterwards, he joined the AEGresearch center in Ulm, Germany, which was latermerged in the DaimlerChrysler corporate research.Johannes is an expert on the the development ofinteractive systems, on the design of user interfaces,and on workflow support. Currently, his research isconcentrating on the integration of distributed E/Edevelopment processes.