Daneva wieringa-camera-ready-re-paper
- 1. Requirements Engineering for Cross-organizational ERP Implementation:
Undocumented Assumptions and Potential Mismatches
Maya Daneva, Roel Wieringa
Department of Computer Science, University of Twente, The Netherlands
m.daneva@utwente.nl, roelw@cs.utwente.nl
Abstract nearly independent business units, each responsible for
their own profit and loss. ERP implementation is
A key issue in Requirements Engineering (RE) for considerably more difficult in such a networked context
Enterprise Resource Planning (ERP) in a cross- than in an intra-organizational context because in a
organizational context is how to find a match between the networked context, we have different business actors who
ERP application modules and requirements for business make decisions based on their own local criteria. Different
coordination. This paper proposes a conceptual businesses have different infrastructures, different
framework for analyzing coordination requirements in enterprise systems, different business processes, different
inter-organizational ERP projects from a coordination semantics of data, different authorization hierarchies, and
theory perspective. It considers the undocumented different decision centers. If these businesses decide to
assumptions for coordination that may have significant cooperate for a particular purpose, all these differences
implications for ERP adopting organizations. In addition, still exist and none of the participating businesses will be
we build a library of existing coordination mechanisms prepared to change their infrastructure, business processes,
supported by modern ERP systems, and use it to make a and semantics, just for this particular cooperation, or to
proposal for how to improve the match between ERP reveal the confidential business rules embedded in their
implementations and supported business coordination processes and applications.
processes. We discuss the implications of our framework ERP implementation is the customization and
for practicing requirements engineers. Our framework and introduction of an ERP system in a (possibly networked)
library are based on a literature survey and the business. One of the most crucial tasks in such a project is
experience with ERP implementation of one of us requirements engineering, in which the properties of the
(Daneva). In future empirical research we will further ERP system to be implemented are aligned to the
validate and refine our framework. requirements of the busines(es) that will use it.
ERP vendors and their consulting partners offer
standard RE processes for ERP projects. Recent research
1. Introduction in ERP RE has identified flaws with these standard
processes and proposed creative solutions to reduce the
From their origin in material requirements planning, cost of ERP RE by avoiding scope creep, involving the
ERP systems have evolved into software packages that right stakeholders, allocating sufficient resources, and
support coordination of different actors in a company. enlisting vendors' consultants'
and support to RE problems
Current ERP systems contain modules not only for [3, 4, 8, 9, 10, 11, 16, 28, 34]. Nevertheless, the central
material management, but also for accounting, human problem of ERP implementation still exists: to find a
resource management and all other functions that support match between the flexibility often required by the
business operations. In the past years, the role of ERP business, and the rigidity usually imposed by the ERP
systems as coordination support has been extended to system. This problem is aggravated in a cross-
cross-organizational coordination. By ‘cross- organizational context because, as we will see, the rigidity
organizational’ we mean that the ERP system is used by of the ERP system is imposed by built-in assumptions
different independent, or nearly independent, businesses. about business semantics, business processes, business
For example, businesses cooperate with their customers, communication channels and business goals. If these
suppliers, and other stakeholders to form value webs. hidden assumptions do not match the business, the
Large companies have structured themselves as sets of business will experience the ERP system as being rigid and
1
- 2. unable to meet the business requirements. In a networked restrictiveness, and (4) package domain specifics versus
context, there is a mismatch between the ERP and each of organization domain specifics. However, their typology
the participating businesses. was empirically derived from cases in an inter-
The present paper proposes to tackle this problem from departmental integration project in one organization in the
the point of view of coordination theory [2,6,29,32,33]. public sector. We explore these opposing forces in the area
Since ERP systems are coordination support systems, we of cross-organizational ERP implementation.
should be able to identify the coordination mechanisms The theoretical perspective that we found most helpful
supported by an ERP system. If we explicitly specify these in examining the issue of misalignments and their sources
mechanisms in a cross-organizational setting, then the was from the field of coordination [29]. Our point of
requirements engineer should be able to find a match departure is the observation that modern ERP systems are
between the coordination support offered by the ERP widely used as an administrative framework for planning,
system and the coordination mechanisms selected by the conducting and monitoring a large array of functionally
cooperating companies. segmented operations in ways that both (i) accommodate,
We will present an inventory of coordination in real time, the intrinsic cross-organizational
mechanisms implicitly assumed by ERP systems, and interdependencies underlying these operations, and (ii)
analyze the role that selection of these mechanisms plays enable their control. Consequently, from coordination
in balancing rigidity imposed by an ERP system against theory perspective, these systems can be viewed as
the flexibility required by the cooperating organizations. coordination technology.
We will see that rigidity will allow the benefits of cross- We follow Malone and Crowston [29] in defining
organizational cooperation to be reaped, whereas coordination as the management of shared actions by
flexibility will decrease the benefits and, at the same time, different business actors. Classic coordination mechanisms
increase the cost of implementing and maintaining the ERP distinguished in economic sociology include market-based
system. coordination, in which goods and services are exchanges
The paper is organized as follows: Section 2 provides based on price, and relational coordination, in which actors
background and related work. Section 3 presents our work towards a common goal based on shared and implicit
inventory of cross-organizational coordination mechanisms norms of behavior [2, 32, 33]. IT mechanisms to support
supported by ERP packages. Section 4 analyzes the coordination include, among others, shared ERP systems
mismatch between business flexibility and ERP rigidity and Enterprise Application Integration middleware [30].
and explains the impact of the coordination mechanisms on The question investigated in this paper is which ERP
this problem bundle. Section 5 discusses the implications mechanisms are available to support different coordination
of our framework for ERP RE and offers some hypotheses mechanisms. The relevance for cross-organizational RE is
that we think are worth further research efforts. Section 6 that an analysis of the desired coordination mechanisms of
concludes with plans for future research. a set of organizations will lead to requirements for ERP
package implementation. Preferences about coordination
2. Background and Related Work mechanisms are usually hidden in the ERP packages and
therefore lead to unpleasant surprises when the chosen
This paper rests on our previous work in ERP RE ERP package turns out not to match at all the implicitly
[9,10,11] and the experiences published by [3,4,8,28,34]. desired but undocumented coordination mechanisms.
Our research also builds on the works of Davenport To the best of our knowledge, so far there has been no
[12,13,14], Hong and Kim [22], Gattiket and Goodhue systematic analysis of the role of coordination
[18], Markus [5,31], Scott & Kaindl [37], and Soh et all requirements in ERP projects. There is to date no unified
[39,40] in analyzing the alignment perspective in ERP framework for describing the various kinds of coordination
projects. Although the ERP literature highlighs the issue of mechanisms, nor a systematic set of rules for dealing with
misalignments in ERP projects, the mere assertion that the coordination needs of organizations. Requirements
they arise from unmet organizational requirements masks engineers and business representatives still have to rely on
the variety of sources of misalignments. Prior studies their intuition and experience, and the problem of
investigating the nature of the ERP mismatches are scarce. coordination is still being confronted in a largely ad-hoc
One exception is the misalignment typology suggested by fashion. Moreover, the few ERP publications that include
Soh et all [40] who adopt a dialectic perspective and coordination aspects in the assessments of ERP systems
extend their initial mismatch categorization [39] by [17,41], describe these aspects only in general terms,
considering four pairs of opposing forces: (1) push towards without characterizing in detail differences between (i)
integration versus differentiation, (2) process orientation how agreements on joint actions are achieved, and (ii) how
versus functional orientation, (3) flexibility versus the default coordination mechanisms in ERP address
2
- 3. those needs. This vagueness makes it difficult to determine the other hand, have built-in assumptions about
what alternative coordination mechanisms might be useful coordination processes at the business level, and this is the
in a given organizational context or to directly translate topic of this paper. Note incidentally that at the enterprise
these alternative coordination process designs into system level, ERP is but one of different possible
specifications of individual activities or uses of ERP to integration technologies such as shared databases, data
support a process (e.g., as part of a business process warehouses, ERP modules and workflow management
redesign effort [13,14,20]). systems [30]. Although these technologies can be provided
together by one ERP vendor, a networked business may
3. An Inventory of Coordination Mechanisms decide to use any multi-vendor combination of them and
we consider them as distinct technologies. However, we
We classify coordination mechanisms based on the hypothesize that at least some of our findings can be
scheme shown in Figure 1. generalized to the other integration technologies at the
enterprise system level.
Our review has yielded the following coordination
mechanisms at the business level:
• Goal-oriented mechanisms referring to the
partners’ agreements on the goals of coordination.
• Process-oriented mechanisms concerned with
establishing end-to-end inter-organizational
processes, for example, client order fulfillment
Figure 1. The framework for business-IT processes or product provisioning processes.
alignment. • Semantics-oriented mechanisms referring to the
partners’ agreements on the definition and the use
The horizontal layers classify entities in a service of common meanings of key information entities.
provisioning hierarchy: physical entities provide services • Communication-oriented mechanisms including
to a software infrastructure, which provides services to the transmission and interpretation of information
enterprise systems, which provide services to businesses. in the networked organization.
We take four views on businesses: Businesses have goals, Table 1 describes which coordination mechanisms our
they perform processes, they communicate with one study covers and how these are supported in state-of-the-
another, and while doing that, they exchange data with art ERP systems. The table structures ERP support for
semantics. This framework is taken from our previous coordination in a networked context and provides
research on business-IT architecture alignment. For examples. It is meant to illustrate the different coordination
motivations we refer the reader to those papers [15,43]. mechanisms and is in no way exhaustive. Indeed, it can’t
Our interest is in the upper two layers of the framework be exhaustive, as new approaches to cross-organizational
because this is where the process and systems alignment in coordination are getting implemented on ongoing basis.
networked organizations takes place. A review of the The crucial observation to make of these mechanisms is
literature on ERP implementation [5,7,12,13,14,23,25, that each one starts with the word ' '
shared' Now, inside
'
.
31,36,38,41] and of our own experience in implementing one company it may often be true that these mechanisms
ERP solutions based on the SAP package [9,10,11] reveals are shared without everyone ever talking about them
that there is a small number of coordination technologies explicitly. After all, within one company we can assume
in use at the enterprise systems level: that there is one culture and one shared way of doing
• Shared database, things. But across different companies, what members of
• Data warehouse, one business silently assume to be the normal way of
• ERP functional application modules, and working can be quite different from what people in another
• Workflow management systems. company silently assume about the normal way of
There are additional infrastructure level technologies working. And even within one company there may be
such as Enterprise Application Integration middleware, severe mismatches, for example if the company consists of
mobile technologies for information sharing, and, still different nearly independent business units. Table 1
experimentally, web service technologies. We surmise that therefore represents a list of hidden assumptions that must
there is little connection between coordination processes at be brought out in the open in cross-organizational ERP
the business level and integration technologies at the implementation. The assumptions about shared
infrastructure levels, and we will not pursue this here. coordination mechanisms may be quite different across
Integration technologies at the enterprise system level, on different business partners.
3
- 4. Types of Implicitly assumed
What it means to ERP adopters? Examples
mechanisms coordination mechanisms
Shared vision of the Unified brand management by using common order
Goal-oriented Presenting one face to clients and
networked organization management, sales force, service & marketing
mechanisms sharing corporate identity
[13,14] analytics applications.
Updating supply chain partner’s databases depends
Shared view of services Motivating dependencies between
on a single event, e.g. when inventory is depleted to
offered by network to clients services of different businesses
a critical amount.
Process- Common agreement about Standardized operational
oriented business process environment procedures, access permissions, and Common payment processing procedures
mechanisms [12,13,14] control patterns
Reducing organizational operations
to a large series of procedural steps
Agreement about process-
tied together to sequences, sub- Creating vendor master files or charts of accounts
orientation [12,13,14]
functional categories, modules and
cross-modular operations
Common agreement on Sharing enforceable business rules
Rules for tracking employee attendance and absence
mngnt policies [13] that are explicit and consistent
Descriptions of the most important
business processes within an
Branch-specific solution maps, like mySAP
Solution maps [23] industry sector, the technologies
Aerospace & Defense solution map [23]
(ERP elements, add-ons) & services
needed to support the processes
Shared understanding of the Trading partner portals, auction and exchange
Shared transaction processing
position of ERP in the cross- mechanisms, catalogs, as provided by Oracle
engines [12]
organizational architecture Exchange [12] and SAP [23]
Semantics-
Shared data dictionary Common definitions of information Maintaining a centralized view of company’s
oriented
[7,9,13,25] entities customers and business partners data
mechanisms
Integrating global data on site capacity, production
Shared reporting formats and Standard presentation formats and
and transportation costs, tariffs, and demand to
semantics [13] information content of output
schedule across multiple sites [13]
Use of self-services to balance top-down control
Delegation about data access Distributed access to data and
with bottom-up empowerment through open
permission [38] distributed application logic
information sharing culture
Common principles of cross-
Data consistency [24] and Data ownership, data modularity, trust: no need of
org. data management
alignment with businesses alternative sources to verify data accuracy [24]
[14,23]
Representing practices embedded in
Reference models
the package in the form of reusable The R/3 Reference Model [7,36]
[7,23,25,38]
process and data models
Configurable master lists to allow product
Industry-specific solution aspects of
Shared product models [25] specification with variances, a requirement specific
the ERP package
to the metal, paper and textile industries
Communicati Agreements about Shared understanding about ERP-package compliant XML schemas, by which e-
on-oriented communication channels and information transmission and network transactions are structured, e.g. RosettaNet
mechanisms standards (at business level) interpretation for electronic industry and Acord for insurance [12]
Bringing information together based on user’s role
within the company (like SAP role-specific portals),
Bringing people to the required
or calling up customers’ purchase history with the
Sharing of knowledge level of understanding to get their
firm, external reports and discussion items from
job done
other sales and service staff members who have dealt
with the customers [12]
Table 1 An inventory of cross-organizational coordination mechanisms.
Furthermore, if we compare the coordination with these authors that a coordination mechanism can
support provided by today’s ERP systems as indicated be characterized by the extent to which it is suited to
in Table 1 to the general coordination mechanisms as different organizational tasks, corporate cultures and
studied in coordination science [29], it becomes clear environments. Thus, in case of cross-organizational
that the latter are too general to be helpful in making ERP projects, coordination mechanisms vary in the
choices in ERP implementations. However, we agree degree to which coordination is prescribed at the time
4
- 5. of RE, the cost in terms of time and effort associated made between rigidity and flexibility in ERP solution
with setting up the mechanism in question, and the design. To check if our library of coordination
degree of change this mechanism brings to the mechanisms can help making those tradeoffs, we
organization at the post-implementation stage. investigated how the four groups of mechanisms fit into
Further research of the coordination perspective in the multifaceted problem description in Figure 3.
the field of ERP may indicate the need for more or
finer distinctions. However, for the time being, we do
benefits
believe that this inventory is adequate for +
understanding the choices for coordination mechanisms
that any ERP adopter is confronted with. _
costs sharing
4. Causal Analysis +
resistance
In this section, we analyze the role that the
coordination mechanisms from Table 1 play in the
clash between flexibility and rigidity typical for ERP Figure 2. High-level representation of
implementation. We explored this link as documented dependencies in ERP implementations.
in the literature [1, 5, 9, 10, 11, 12, 13, 14, 17, 18, 19,
20, 21, 22, 24, 25, 26, 27, 31, 35, 36, 37, 38, 39, 40, We made the following observations: First, Figure
41, 42] in order to understand which problems and 3 not only expands the four boxes of Figure 2 but it
misalignments could have been spotted and understood adds new insights into how benefits are shifting in
early in the ERP project, if the undocumented dependence on the coordination decisions that may
assumptions would have been brought up as part of favor either process standardization (and rigidity of the
ERP RE and coordination requirements would have ERP solution) or process diversity (and flexibility of
been specified as part of the business requirements. the solution). For example, the more diverse the ERP
The review of our above cited sources let us derive adopters decide to keep their processes (and the more
the detailed problem dependency map (in both intra- flexible the solution they want to design), the more the
and inter-organizational context) shown in Figure 3. As options for fostering creativity and maintaining the
an introduction to it, we first present a high-level spirit of innovation in the organization [24]. The latter
summary of the map in Figure 2. (In preparing this two benefits, however, will be of less consideration, if
paper, we derived Figure 2 as an abstraction of Figure the ERP adopters decide on a higher level of process
3.) Figure 2 says that integration benefits increase standardization. This decision, in turn, will favor the
through more sharing, e.g. sharing of standardized and realization of the benefits due to sharing, namely,
harmonized processes and common data. Also, more reduced transaction costs, organizational transparency,
sharing decreases the total costs of ownership. On the data visibility, data accuracy.
other hand, the more the organizational processes get Second, Figure 3 reflects the fact that benefits from
integrated via the shared process and data environment, bringing ERP in never come cheaply. It makes it clear
the more these get adapted to the default ERP what price ERP adopters should expect to pay in order
structures, and so the more the change imposed on the to realize the benefits due to sharing or the ones due to
organization and the more the organizational resistance flexibility. Getting a more flexible solution means
to it. We believe Figure 2 maps the basic problem of customizing the system to fit the business processes,
rigidity versus flexibility in ERP implementation. which also means that cost, like customization (29),
Detailed analysis of the reported experiences maintenance (30) and testing costs (33), and risks, like
brought us to the multifaceted representation of the customization (27), system performance (31) and
ERP problem space in Figure 3, in which the boxes release lag risks (26), will increase. On the other side,
represent typical issues that adopters encounter or try opting for more sharing means incrementally or
to avoid in ERP implementations, and the directed radically changing (cross-) organizational business
arrows show causality. The references that suggest the processes to fit the system. It means less customization
33 links presented here are provided as appendix to this and maintenance costs as well. The price for
paper. Figure 3 shows generic domain knowledge about implementing these changes, though, comes in the form
the impact of choices of an ERP system. It reflects the of costs for managing and coordinating large-scale
opposing forces of process standardization and process business process changes and coping with politics,
diversity [20] and, also, shows that tradeoffs need to be resistance and corporate inertia [5,39,40].
5
- 6. Figure 3. The problem dependency map in ERP Implementation practice.
Third, as the coordination mechanisms from Table 1 companies have to issue and receive transactions
are available for ERP clients to achieve sharing, they directly to and from their ERP systems without any
clearly tend to support rigidity, reuse (10), human intervention [12]. This requires changes in how
standardization (12), and integration (13). These things get done internally in each of the partner’s
observations and the fact that the coordination organizations (16), as handling business transactions is
mechanisms from our library are all about sharing no longer limited by organizational boundaries. This
encouraged us to consider substituting the boxes in the would most likely be in opposition to other forces
sharing-labeled gray area in Figure 3 with the groups arising from each partner’s organization having their
of coordination mechanisms from Table 1. Indeed, if own systems.
we replace boxes 10, 12, and 13 in Figure 3 with the Process-oriented mechanisms. A process
groups of coordination mechanisms from our library, orientation in an inter-organizational ERP context
one can clearly see how our library fits in and what means predisposing each company to manage itself
type of problems it could potentially help to explain along business processes. Also, it means shared process
(Figure 4): ownership. This requires redistribution of management
Goal-oriented mechanisms. An orientation towards responsibility in each of the partner companies as
shared goals pushes the partners in the networked shared process ownership can’t be imposed on a
organization towards (i) developing a clear sense of fragmented organization [20]. Differences between new
who and what they are, and (ii) certainty as to how the agreements on shared processes and previously used
network delivers value to customers and how it business practices typically lead to disruption to
differentiates itself. In order for networked business (18), and, ultimately, increased resistance to
organizations to be economically advantageous, partner change (19).
6
- 7. Figure 4. The problem dependency map: a coordination perspective.
Semantics-oriented mechanisms. Next, a change as business opportunities arise, are taken
semantics orientation implied through the coordination advantage of, then abandoned as their value diminishes.
mechanisms pushes the partner companies towards The above observations give an indication that the
adopting a common terminology for those areas of library of coordination mechanisms has its
business activity in which the partner want to do things corresponding components in the problem dependency
together. This gives rise to cross-organizational issues map (in Figures 3 and 4). Thus, our preliminary
in terms of common data structures, data ownership analysis allows us to draw two early conclusions: First,
and mastering, data flow transparency, responsibility exploring coordination requirements by addressing the
for data entry and updates, and related changes in 15 coordination mechanisms in our library (Table 1)
workflow [40]. can be an important step towards unpacking the
Communication-oriented mechanisms. Finally, dimensions on which misalignments can arise. Second,
communication-oriented coordination mechanisms tend the library of mechanisms can serve as a preliminary
to support interoperability standards that directly inventory on which critical inter-organizational
contribute to building “collaborative communities” integration issues, costs and risks can be surfaced early
[12]. The underlying assumption is that partner in the ERP project.
companies have to realign horizontally and the variety
of shared tasks that are performed requires less 5. Conclusions and Further Research
flexibility. In networked context, however, this
assumption contradicts with the need of each of the In this paper, we attempted to show what the role of
partner organizations to dynamically build connections the undocumented built-in ERP assumptions is in inter-
that handle specific portion of the shared process, that organizational ERP RE. We took an inventory of
7
- 8. existing coordination mechanisms and mapped them addressed by a networked business suggests alternative
onto typically encountered problems identified in coordination mechanisms that could be used, thus
empirical studies. We presented a perspective that, we creating a space of possible business process designs.
believe, helps the requirements engineers to develop an
understanding of the opportunities and issues 6. Acknowledgement
associated with the ERP coordination mechanisms as The authors would like to thank Dr. Pascal van Eck and
undocumented assumptions: First, our problem the anonymous reviewers for their comments and
dependency map is a problem domain theory; it allows suggestions that led to the improved version of our
the requirements engineers to reason about the impact paper.
of choices. Second, the undocumented assumptions
make the coordination choices more explicit. Our
library not only can facilitate interdisciplinary transfer 7. References
of knowledge about ERP-supported coordination, it [1] Al-Mashari, M., Process Orientation through Enterprise
provides a guide for analyzing organization-specific Resource Planning: A Review of Critical Issues, Knowledge
and Process Management, 8(3), 2001, pp. 175-185.
coordination needs and generating alternative ways to
[2] Alstyne, M., The State of Network Organizations: A
fulfilling them. The variety of coordination Survey in Three Frameworks, Journal of Organizational
mechanisms that we analyzed and included in our Computing, 7(3), 1997, pp.83-151.
library is not found in previous research. Also, we [3] Alvares, R., J. Urla, Tell Me a Good Story: Using
provided a start in the direction of how to organize Narrative Analysis to Examine Information Requirements
these coordination mechanisms. In addition, we used Interviews during an ERP Implementation, The DATA
real life examples to motivate our analyses. BASE for Advances in IS, 33(1), pp. 38-52
Each directed arrow in our problem dependency [4] Arinze, B., M. Anandarajan, A Framework for Using OO
Mapping Methods to Rapidly Configure ERP Systems,
map represents a hypothesis that can become a subject
Comm. of ACM, 46(2) pp. 61-65.
of future empirical validation studies. Thus, for IS [5] Brehm, L, A. Heinzl, M.L. Markus, Tailoring the ERP
scientists, we formulated 33 hypotheses with a very Systems: a Spectrum of Choices and Their Implications,
preliminary analysis that indicates that it will be useful Proc. of the Hawaii Conf.on Systems Sciences, 2004.
to do this research. [6] Clemons E.K., S. P. Reddi, M.C. Rows, The Impact of
We believe that our approach provides a meaningful Information Technology on the Organization of Economic
starting point in classifying ERP misalignments. Activity: The `Move to the middle'Hypothesis, Journal of
` '
However, for the framework to be useful at application Management Information Systems, 10(2), 1993, pp. 9-35.
and project level in the long-term and to progress from [7] Curran, T., A. Ladd, SAP R/3 Business Blueprint, 2nd ed.,
Prentice Hall, 2000.
its current state, more analytic capabilities need to be
[8] Dalal, N.P., M. Kamath, W.J. Kolarik, E. Sivaraman,
built-in. Therefore, our immediate plans are to use it as Towards an Integrated Framework for Modelling Enterprise
a vehicle to explain typical misalignment phenomena in Processes, Comm. of ACM, 47(3) 2004, pp. 83-87.
cross-organizational implementations and to refine it [9] Daneva, M., ERP Requirements Engineering Practice:
based on experiences we will collect in case studies. Lessons Learnt, IEEE Software, 21(2), 2004, pp. 26-33.
As our proposal rests on cases from the ERP [10] Daneva, M., Patterns of Success and Failure in ERP
implementation practice, we are interested in knowing Requirements Engineering Proc. 12th Int’l Workshop on
if our ideas can be extended to projects implementing Software Metrics, Shaker, Aachen, 2004, pp. 527-546.
other technologies for inter-organizational integration, [11] Daneva, Using Maturity Assessments to Understand the
ERP Requirements Engineering Process, Proc. Joint Int’l
like data warehouses, workflow management systems,
Requirements Engineering Conf., IEEE CS Press, 2002,
or Enterprise Application Integration middleware [30]. pp.255–262.
This will be subject to validation in field research too. [12] Davenport, T., The Future of Enterprise System-Enabled
Given ERP coordination mechanisms support a Organizations, Information Systems Frontiers 2(2), 2000, pp.
variety of intra- and inter-organizational interactions 163-180.
[17], to design a new RE process for cross- [13] Davenport, T., Mission Critical: Realizing the Promise
organizational ERP implementations, it will be useful of Enterprise Systems, HBS Press, 2000.
to consider alternative coordination mechanisms that [14] Davenport, T., Putting the Enterprise into the Enterprise
could be used to manage data and process sharing. One System, Harvard Business Review, 76(4), 1998, pp. 121-131.
[15] Eck, P.A.T. van, H. Blanken and R.J. Wieringa, Project
question that comes out of this paper and, we think,
GRAAL: Towards Operational Architecture Guidelines, IJ of
seems worth exploring, is: In what ways an ERP system Cooperative Information Systems, 13(3), 2004, pp. 235-255.
can be arranged differently while achieving the same [16] Esteves, J., J.A. Pastor: Establishing the Importance of
goals? Understanding the coordination problems ERP Implementation - Critical Success Factors along ASAP
8
- 9. Methodology Processes, Proc. of Int’l. Conf. on Enterprise [30] Markus, M. L., Paradigm Shifts - Business and Business
Information Systems, 2001, pp. 182-187. /Systems Integration, CAIS 4 (10) 2000.
[17] Fan, M., N. Stallaert, A.B. Whinston, The Adoption and [31] Markus, M.L., C. Tanis, P.C. v. Fenema, Multisite ERP
Design Methodologies of Component-based Enterprise Implementations, Comm. of ACM, 43(4), 2000, pp. 42-87.
Systems, European Journal of IS, 9, 2000, pp. 25-25. [32] Miles, R.E. and C.C. Snow, Causes of Failure in
[18] Gattiker, T.F., D.L. Goodhue, Understanding the Local- Network Organizations, California Management Review,
level Costs and Benefits of ERP through Organizational 34(4) 1992, pp. 53-72.
Information Processing Theory, Information & Management, [33] Powell, W.W., Neither Market nor Hierarchy: Network
41, 2004, pp. 431-443. Forms of Organization, Research in Organizational Behavior,
[19] Grossman, T., J. Walsh, Avoiding the Pitfalls of ERP 12, 1990, pp. 295-336.
System Implementation, Information Systems Management, [34] Rolland, C., N. Prakash, Bridging the Gap between
Spring 2004, pp. 38-42. Organizatopnal Needs and ERP Functionality, Requirements
[20] Hammer, M., S. Stanton, How Process Enterprises Engineering, 5, 2000, pp. 180-193.
Really Work, Harvard Business Review, 77(6), 1999, pp. [35] Ross, J. W., M. R. Vitale, The ERP Revolution:
108-118. Surviving vs. Thriving, Information Systems Frontiers 2(2),
[21] Hirt, S.G., E.B. Swanson, Emergent Maintenance of 2000, pp. 233-241.
ERP: New Roles and Relationships, Journal of Software [36] Scheer, A.-W., F. Habermann, Making ERP a Success,
Maintenance & Evolution: Research and Practice, 13, 2001, Comm of ACM, 43(4), pp. 57-61.
pp. 373-397. [37] Scott, J., L. Kaindl, Enchancing Functionality in an
[22] Hong, K.-K., Y.-G. Kim, The Critical Success Factors Enterprise Software Package, Information & Management,
for ERP Implementation: an Organizational Fit Perspective, 37, 2000, pp. 111-122.
Information & Management 40, 2000, pp. 25-40. [38] Selchert, M., Enhanced Project Success through SAP
[23] Kagermann, H., G. Keller, MySAP.com Industry Best Practices - International Benchmarking Study, ISBN 1-
Solutions, Addison Wesley, 2000. 59229-031-0, SAP Press, 2004.
[24] Kallinikos, J., Deconstructing Information Packages: [39] Soh C., S.S. Kien, J. Tay-Yap, Cultural Fits and Misfits:
Organizational and Behavioral Implications of ERP Systems, Is ERP a Universal Solution? Comm. of ACM 43(4), 2000,
Information Technology & People, 17(1), 2004, pp. 8-30. pp. 47-51.
[25] Keller G. and T. Teufel, SAP R/3 Process Oriented [40] Soh, C., S.K. Sia, W.F.Boh, and M. Tang,
Implementation, Addison-Wesley, 1998. Misalignments in ERP Implementations: a Dialectic
[26] Luo W., D.M. Strong, A Framework for Evaluating ERP Perspective, IJ HCI, 16(1)2003, pp. 81-100.
Implementation Choices, IEEE Trans. On Engineering [41] Stefanou, C.J., A Framework for the Ex-ante Evaluation
Management, 51(3) 2004, pp.322-333. of ERP Software, European Journal of IS, Special Issue on
[27] Madapusi, A., D. D’Souza, Aligning ERP Systems with IT Evaluation, 10(1) 2001, pp.204-215.
International Strategies, Information Systems Management, [42] Stevens, C., Enterprise Resource Planning: a Trio of
Winter 2005, pp. 7-17. Resources, Information Systems Management, Summer
[28] Maiden N.A., C. Ncube, “Acquiring COTS Software 2003, pp. 61-67.
Selection Requirements,” IEEE Software, 15(3), 1998, pp. [43] Wieringa R.J., H.M. Blanken, M.M. Fokkinga, and
46–56. P.W.P.J. Grefen, Aligning Application Architecture to the
[29] Malone, T., K. Crowston, The Interdisciplinary Study of Business Context, Proc. Of Conference on Advanced
Coordination, ACM Computing Surveys, 26(1), 1994, pp. Information System Engineering (CAiSE 03), LNCS 2681,
87-119. Springer, 2003, pp. 209-225.
9
- 10. 8. Appendix
Link Proposition and Relevant References
1 The tighter the integration of operational and informational procedures, the better the compliancy with regulatory
requirements [31,39,40].
2 The tighter the integration of operational and informational procedures, the lower the transaction costs [12,13,14,20].
3 The tighter the integration of operational and informational procedures, the better the coordination across functions
and sites [1,12,13,14,31,39].
4 The tighter the integration of operational and informational procedures, the more transparent the organization
[12,13,14,39,40].
5 The tighter the integration of operational and informational procedures, the better the ability of the organization to
handle data complexity [12,13,14,24,25].
6 The tighter the integration of operational and informational procedures, the higher the level of data accuracy that the
organization can achieve [12,13,14].
7 The tighter the integration of operational and informational procedures, the better the level of data visibility that the
organization can achieve [12,13,14,39,40].
8 The more rigid the solution, the higher the costs of building consensus among stakeholders [24].
9 The more rigid the solution, the higher the reuse risks [9,10,11].
10 The more rigid the solution, the higher the levels of reuse an organization can achieve [7,9,10,11,25].
11 The more rigid the solution, the stronger the push towards inter- and intra-organizational integration
[5,12,13,14,20,31,35,39,40].
12 The more rigid the solution, the more standardized the business processes [5,9,12,13,14,17,20,24,25,26,27,36,39,
40,42].
13 The more rigid the solution, the tighter the integration of operational and informational procedures that the
organization can achieve [1,5,9,10,11,12,13,14, 17, 18,19,20,21,22,24,25,26,27,31,35,36,37,38,39,40,41,42].
14 The more rigid the solution, the better the level of interoperability that can be achieved [1,24,25].
15 The more rigid the solution, the more predictable the global business processes [12,13,14,20,25].
16 The tighter the integration of operational and informational procedures, the greater the changes in processes and data
management imposed on the organization [1,5,9,10,11,12,13,14,17,18,19,20,21,22,24,25,26,27,31,35,36,37,38,39,
40,41,42].
17 The tighter the integration in terms operational and informational procedures, the more complex the changeover
management processes [5,13].
18 The greater the changes in processes and data management imposed on the organization, the greater the disruption to
business [12,13,14,20,31,39,40].
19 The greater the changes in processes and data management imposed to the organization, the more organizational
resistance to them and the more the potential sources for political issues [1,12,13,14,19,20,31,39,40].
20 The more predictable the global business processes, the less the cycle time and the better control over cycle times
[12,13,14,20,25].
21 The more rigid the solution, the less flexibility it offers to business users [1,5,9,10,11,12,13,14,17,18,19,20,
21,22,24,25,26,27,31,35,36,37,38,39,40,41,42].
22 The more flexible the solution, the more the options for fostering innovative thinking [22,24].
23 The more flexible the solution, the more the options for inventing creative ways of working [22,24].
24 The more flexible the solution, the more compliant it is with the specifics of the organization [39,40].
25 The more flexible the solution, the more diverse the organizational business processes [1,5,9,10,11,12,13,14,17,18,19,
20,21,22,24, 25,26,27,31,35,36,37,38,39,40,41,42].
26 The more flexible the solution, the higher the risks of release lags [21,41].
27 The more flexible the solution, the higher the customization risks [5,12,13,14,26,27,31].
28 The more flexible the solution, the more the upgrade difficulties [5,12,13,14,26,31].
29 The more flexible the solution, the higher the customization costs [9,12,13,14,21].
30 The more flexible the solution, the higher the maintenance costs [9,12,13,14,21].
31 The more flexible the solution, the higher the system performance risks [39,40].
32 The more flexible the solution, the more complex the user domain (e.g. the more the data views that need to be
consolidated, the more the interfaces that need to be maintained) [9,12,13,14,21].
33 The bigger the scope of customization, the more the testing efforts that are required [9,12,13,14,21].
10