See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/224080013

Estimating the Benefits and Risks of Implementing E-Procurement

Article  in  IEEE Transactions on Engineering Management · June 2010

DOI: 10.1109/TEM.2009.2033046 · Source: IEEE Xplore

CITATIONS READS
54 5,992

2 authors:

Peter Trkman Kevin Mccormack

University of Ljubljana Northwood University
66 PUBLICATIONS   3,585 CITATIONS    56 PUBLICATIONS   2,731 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Business analytics and business value View project

Practitioner oriented presentation of my research View project

All content following this page was uploaded by Peter Trkman on 16 May 2014.

The user has requested enhancement of the downloaded file.

338 IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT, VOL. 57, NO. 2, MAY 2010

Estimating the Benefits and Risks of Implementing

E-Procurement
Peter Trkman and Kevin McCormack

Abstract—In recent years, organizations have invested heavily on real-life data. The main methodological approach is the sim-
in e-procurement technology solutions. However, an estimation of ulation of business processes since the value of information
the value of the technology-enabled procurement process is often technology (IT) implementation should be measured at the pro-
lacking. Our paper presents a rigorous methodological approach
to the analysis of e-procurement benefits. Business process simu- cess level [8]. The approach also enables the recommendation
lations are used to analyze the benefits of both technological and of changes to an organizational structure (e.g., the appropriate
organizational changes related to e-procurement. The approach approval policy). In addition, risks are estimated based on a
enables an estimation of both the average and variability of pro- novel application of the value-at-risk (VaR) concept. The pre-
curement costs and benefits, workload, and lead times. In addition, sented calculations can be easily replicated in different practical
the approach enables optimization of a procurement strategy (e.g.,
approval levels). Finally, an innovative approach to estimation of settings with different input data.
value at risk is shown. The structure of the paper is as follows. First, the main ben-
efits and challenges of e-procurement introduction and usage
Index Terms—Business process modeling, e-procurement,
methodological approach, procure-to-pay (P2P) process, simula- are summarized. Second, the main objectives of the study are
tions, value at risk (VaR). outlined. The methodology, models, and collected data are then
presented, followed by the results and analysis. Finally, the
I. INTRODUCTION main implications and limitations of the proposed approach are
HE procure-to-pay (“P2P”) process has become a major outlined.
T challenge to companies that have adopted global sourcing
and distribution as a strategic component of their business. In
recent years, organizations have adopted management practices II. E-PROCUREMENT: BENEFITS AND CHALLENGES
and technologies designed to reduce transaction costs (TCs) [1], Process efficiency and process integration capabilities of a
which includes the automation of P2P process. E-procurement is procurement process provide a significant contribution to firm
the use of electronic means (the Internet, Web, e-mail) to enable performance [9]. The main benefits of e-procurement are an
purchases of products and services over the Internet [2], [3]. It is increase in firms’ competitiveness through cost reduction and/or
believed that, in addition to a decrease in costs, e-procurement boosted efficiency with inbound logistics [10]. These benefits
also eliminates paperwork, improves data accuracy, collabora- can materialize in a reduction of purchasing transactions costs,
tion, and transparency of the process when reducing inventory order fulfillment and cycle time, a reduction of the number of
levels and lead times [4], [5]. suppliers or even a reduction in the price paid, and the number
The challenge is how to measure the increase in efficiency of staff to support purchase transactions [7]. However, in order
(both value and risks) of e-procurement implementations and si- to reap the full benefits, the business processes connected to
multaneous changes in the organization and strategy. Although procurement should be carefully analyzed and (if necessary)
various research and practitioner papers have dealt with the improved before it is supported with an e-procurement solution
question about how to estimate the benefits of e-procurement [11]. In less process mature companies, ordering and receiving
implementation (e.g., [6] and [7]), an answer has been mainly are not connected, and this results to extensive manual matching
offered based on either rough estimates without explaining the and resolution before payment. Information systems are manual
exact methodology or self-reported data from the studied com- and decentralized, while information resides on spreadsheets in
panies. Both approaches have issues when estimating the poten- individual computers [6].
tial benefits in a particular company. A key business process impacted by e-procurement is the
Therefore, the main contribution of the approach presented P2P process that encompasses activities from need specifica-
in this paper is that it enables an estimation of savings based tion, sourcing decision, contract-/purchase-order generation, re-
ceipt of material/documents, and finally, settlement and pay-
Manuscript received June 21, 2008; revised October 9, 2008, March 10, 2009, ment. Therefore, the paper’s focus is not on the procurement
and June 23, 2009. First published November 13, 2009; current version pub-
lished April 21, 2010. Review of this manuscript was arranged by Department department, but on the whole procurement process (as shown in
Editor T. Ravichandran. Fig. 1).
P. Trkman is with the Faculty of Economics, University of Ljubljana, Companies are increasingly considering procurement as a
Kardeljeve pl. 17, 1000 Ljubljana, Slovenia (e-mail: peter.trkman@ef.uni-lj.si).
K. McCormack is with the DRK Research—A Practitioner Oriented Re- strategic-level concern of developing a competitive advantage
search Organization, Raleigh, NC 27526-8484 USA (e-mail: kmccormack@ [12]. Lower information exchange costs coupled with lower TCs
drkresearch.org). can also make bilateral relationships more efficient and interfirm
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. operations better coordinated, a phenomenon characterized as
Digital Object Identifier 10.1109/TEM.2009.2033046 the integration effect of IT [13]. IT can reduce the overall TC
0018-9391/$26.00 © 2009 IEEE

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
TRKMAN AND MCCORMACK: ESTIMATING THE BENEFITS AND RISKS OF IMPLEMENTING E-PROCUREMENT 339

Fig. 1. Overview of the procure to pay process.

and risks associated with obtaining goods and services from line with earlier research, which found that early adopters em-
suppliers [14], while also reducing the search costs connected phasize cost reductions and administrative efficiencies from e-
with procurement importantly [15]. procurement [24]. More mature users focus on strategic ad-
Our study includes the whole P2P process, since not only vantage and generate this through organizational changes. E-
order and delivery are vital, but also the activities such as re- procurement impacts both a firm’s primary business processes
ceipt of invoice, approval, payment, and reconciliation can carry and the organizational structures used to coordinate these pro-
considerable costs and/or risks [16]. Past experience has shown cesses [25].
that many efforts have failed because they targeted processes Accordingly, our analysis centers around two main questions:
contained only in a single department [17]. 1) how can the reduction of procurement cost, lead times,
Most importantly, the benefits, costs, and risks of IT im- and employee workloads be measured and
plementation need to be identified, managed, and controlled if 2) which advantages and potential risks do organizational
businesses are to derive value from their investments [18]. De- changes (in our case, a change in approval procedures)
spite the greater attention paid to the P2P, most companies are bring to the procurement process?
still unsure of the benefits and the ways to measure the value In relation to 1), coordination costs, such as search, ne-
of the informatization of procurement, and which factors affect gotiation, communication, follow-up, and error reconciliation
this value [19]. In many cases, the real benefits are not identi- with suppliers, can make up a significant part of costs and e-
fied, resulting in companies not recognizing the true value of procurement can play a vital role in reducing such costs [19].
e-procurement [20]. A recent study showed that IT plays a sig- Earlier research found similar estimates of cost reductions due
nificant role in everyday procurement, but the expectations of to the implementation of e-procurement. Various independent
IT are rarely completely fulfilled [21]. studies found that costs of manually processing a purchase
Therefore, there is a need to better understand the value of e- order can range from U.S. dollar (USD) 100 to 250, while
procurement at a level of analysis smaller than a firm [19]. This e-procurement can reduce these costs to around USD 10 to
paper analyzes its value at the level of the business process. The USD 30 [6], [16], [26], [27].
analysis of the intermediate outcome of e-procurement can shed In such a way, e-procurement should lead to savings of around
more light on the efficiency of the process than financial outcome 42%–65% in purchasing TCs [7], [28]. In addition, sourcing cy-
measures, such as return on assets or return on investment [3]. cling times should be reduced by 25%–30% and time to market
This is vital since so far there has been only limited effort to by 10%–15%. The Aberdeen research gave even higher esti-
conceptualize the key constructs that characterize procurement mates, namely a reduction of requisition-to-order cycles by 66%
as a process [12]. While the usage of technology is important, and a reduction of costs by 58% [29].
the process approach allows the identification of key organiza- Obviously, all of these estimations similarly indicate rela-
tional and other issues [11], [22]. In order to realize the benefits tively large savings, usually around 50% or more of total costs
of e-procurement, it is necessary to properly improve the pro- of the procurement process. The only significant exception is
cess and not simply to automate the existing methods of work- reported in [30], who claimed that e-procurement reduced TCs
ing [5]. Then, processes have to be continuously measured and by approximately 99.7%. However, it is likely that such a figure
analyzed by defining and implementing performance measures disregards various costs connected to e-procurement.
and key performance indicators [23]. Therefore, our paper con- Despite the abundance of such studies, most of these es-
ceptualizes P2P as a process and proposes related performance timations either failed to provide a methodology of their ap-
indicators and measurements. proach or used a survey/sample of self-reported benefits from
studied companies for estimating the benefits. This is ques-
III. RESEARCH OBJECTIVES tionable since there is little consensus on how to gauge the
Our approach studies both the prediction of changes due to value of technology-enabled procurement processes [3]. The
technology implementation and organizational savings, and is range of these estimations is too large to be used in a practical

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
340 IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT, VOL. 57, NO. 2, MAY 2010

setting and organization-level factors, such as spend characteris- 2) was a P2P process of six major companies. Each company
tics/portfolio and internal sourcing competency, significantly af- (except one) had at least 40 000 purchase orders passing through
fect the level of savings [31]. Further, the assumption of most of their P2P system every year. These companies were in the oil
the studies is that the company will use e-procurement for all of industry equipment, chemical, cement, and oil exploration and
their purchases; yet, in practice, the level of using e-procurement development industries. General data about the companies in-
can be around 25% of the total value of procurements [7], [32]. cluded are shown in Table I (the company names are fictional,
These findings are difficult to replicate and test in other set- all the other data are real).
tings, and are of questionable validity for use as guidance for In order to map the current business process (AS–IS) (activ-
managers facing the need to estimate the benefits and justify the ity 3), semistructured interviews with approximately 100 people
investment in e-procurement solutions. The justification of the from the six companies were conducted. Various procurement
costs and benefits is needed to secure enterprise funding and and IT employees at all levels of the hierarchy and in differ-
support [31]. Therefore, the initial objective of the paper is to ent geographical locations (field, corporate) participated in the
provide a robust simulation-based methodology for analyzing interviews. The following employees from the studied compa-
the reduction in costs/lead times/employee workload that can be nies were included: buyers, procurement managers, commodity
easily replicated by other companies, considering an investment managers, Directors of procurement, Directors of the supply
in the P2P process. chain, IT managers, Vice President of the supply chain, and
In relation to 2), in addition to savings in procurement costs, strategic sourcing managers. The exact titles vary from com-
e-procurement can change the organizational structure, respon- pany to company; the titles also depend on the maturity of the
sibilities, and internal power structures simultaneously [33], and procurement process in each company. In addition, semistruc-
induce a change in organizational processes and culture [10]. tured phone interviews with account managers from approxi-
E-procurement leads to changes at different levels, including mately 30–50 of the largest suppliers of each of the six studied
organizational, financial, and the information systems depart- companies were conducted.
ment [10]. The structure of buying centers tends to flatten The developed business process model is shown in Fig. 3.
and fewer levels of management actively engage in each activ- While the presented model is a simplified version of reality, it is
ity [34]. E-procurement can lead to long-term efficiency gains by sufficient for our purpose. The developed model only includes
fundamentally changing the coordination mechanics and trans- activities in the order and initiation stage, and not those involved
action practices [3]. A typical example is the automatization of in the search for products/suppliers. IT usage in the order and
the approval process of senior managers by preauthorizing oper- initiation stage, for example, has a more significant impact on
ating personnel [34]. As outlined earlier, these changes can bring procurement-process performance [42]. Thus, the assumption
both benefits and potential risks due to mistakes or fraud by em- (that also matches the scope of the collected data) that the com-
ployees. This highlights the critical need to study risk tradeoffs pany procures from known, long-term suppliers was made (see,
and gauge the business value impacts of potential shocks [35]. e.g., [43]). The process model was validated (activity 4) with
Therefore, the second objective of the paper is to measure both company employees involved in the P2P process. Further refine-
the variability of the process time and costs, and the potential ments were made to assure that the developed model matched
risks of these changes also. The latter is measured by the VaR the real situation in the company.
measure. In order to collect the necessary data for the simulation (ac-
The hesitation to adopt e-procurement, for example, does not tivity 5), each of the interviewees was asked to estimate the
stem from expected difficulty or constraints, but arises due to average lead time/time of each activity, along with the variabil-
being unaware of clear anticipated benefits [36], [37]. There is ity of these times. The data acquired from all the interviews
a positive relationship between beliefs about a target new tech- were compared and refined. They were also cross-checked with
nology, its usefulness, and its subsequent adoption [38], [39]. A the data acquired from the SAP system. Descriptive statistics
company’s vague statement of the benefits, leading to an uncer- (frequency, mean, standard deviation, distributions, etc.) were
tain allocation of responsibility for managing their delivery, is used to examine the types of orders (procurement card (PCard),
the number one cause of project failures [40]. e-catalog, and buyer assisted) for each company’s process. A
However, the benefits are not the only determinants of e- lognormal distribution was used for sampling the times of ac-
procurement usage as perceived risks also play an important tivities because it is a sufficiently flexible theoretical probability
role. Thus, the focus should not only be on benefits for firms, distribution for modeling operation times in procurement [44].
but also on estimating the firms’ risks or at least their perception Descriptive statistics for the simulation study are included in
of risks [41]. Table II and only the data for one company are shown; however,
the data for the others are quite similar.
IV. METHODOLOGY All companies were similar in the activities and duration,
The general approach to such analysis is shown in Fig. 2. since they all had the same enterprise resource planning soft-
The activities in the figure are numbered and referred to in this ware, namely SAP, as their procurement system. Although the
section. A parallel gateway (plus sign) indicates tasks that can use of the same software does not automatically mean the use
be done simultaneously. of same processes or process execution efficiency, in our case,
The main objectives (activity 1 in Fig. 2) of our analysis are their processes were built around the SAP recommendations.
outlined in the previous section. The studied process (activity Enterprise resource planning implementation, for example,

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
TRKMAN AND MCCORMACK: ESTIMATING THE BENEFITS AND RISKS OF IMPLEMENTING E-PROCUREMENT 341

Fig. 2. Flowchart of the project.

TABLE I
CHARACTERISTICS OF THE STUDIED COMPANIES

Fig. 3. P2P business process model.

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
342 IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT, VOL. 57, NO. 2, MAY 2010

TABLE II
LEAD TIME OF THE MAIN ACTIVITIES IN THE P2P PROCESS

TABLE III
AVERAGE LEAD TIME AND TOTAL FTES FOR EACH SCENARIO

requires the reengineering of a business process prior to the 2) Scenario 2 (scen2): 80% of orders are buyer-assisted,
adoption of Enterprise Resource Planning (ERP), and the com- while 10% of orders are made with a Pcard, while for
pany usually accepts the proposed standard business processes 10% of orders an e-catalog is used.
from ERP [45]. Since our business process model (Fig. 3) is 3) Scenario 3 (scen3): As the use of new procurement tech-
intentionally rather general, it can be applied to all studied niques increases, only 60% of orders are buyer-assisted,
companies. If a company would have considerably different while the PCard/e-catalog are used in 30%/10% of cases.
P2P process, the whole approach (shown in Fig. 2) should be 4) Scenario 4 (scen4): The level of automation is the same
repeated. as in scen2 (80%, 10%, 10%). However, due to the em-
Based on data from Judril, the value of each order was es- powerment of the employees and organizational changes,
timated as an exponential distribution with the mean of USD the approval levels are tripled.
2000. The current approval levels are that all orders above USD Discrete-event simulations (activity 7) were used to analyze
1000 have to be approved by the supervisor (approval level 1). different scenarios. The reason is that the adoption of rigorous
All orders above USD 10 000 have to be approved by both the business process simulation methodologies enables one to evalu-
supervisor and the senior manager (approval level 2). ate different configuration of process chains in realistic settings
Different scenarios for process redesign and informatization and estimate the expected payoffs resulting from reengineer-
(activity 6) were prepared. The main objective of improving the ing/IT incorporation [49]–[52].
P2P process was to replace some of the high-cost buyer-assisted The main problem of simulations can be the large costs and
orders by introducing both an electronic catalog (e-catalog) and amount of time needed [53]. It is, therefore, often too expensive
a PCard. for small- and medium-sized companies to build simulation
The e-catalog is one of the most widely used e-procurement competence within the company, especially due to the high
technologies and usually contains specifications and prices of all expenditure on specific know-how [54]. Further, it should not be
products obtained from contracted suppliers. Suppliers can di- forgotten that the models developed are always a simplification
rectly access the enterprise server and update information about of a real system under examination.
their products and services [5], [10], [46]. A PCard is an elec- In our case, the Igrafx Process 2007 was used. It is one of the
tronic transaction card issued at the firm level and intended for most widely used simulation tools [55], which enables our ap-
small value transaction, noninventory/stock, and noncapital pur- proach to be repeated in a practical setting. Similar methodolog-
chases. The advantage of PCards is the ease with which they can ical approaches were successfully used in the past to measure
be implemented and the low initial cash investment [7], [47]. the effects of a business process improvement in public adminis-
The analysis examines the following four scenarios (they were tration [56], supply chains [11], and production processes [57].
chosen since they are the most typical situations encountered in A 12-month simulation was run, which amounted to approxi-
the real world [32], [48]). mately 33 000 orders. The time between each transaction (order)
1) Scenario 1 (scen1): All orders (100%) are buyer-assisted. is randomly distributed. In addition to process models and data

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
TRKMAN AND MCCORMACK: ESTIMATING THE BENEFITS AND RISKS OF IMPLEMENTING E-PROCUREMENT 343

from each of the four scenarios, the following input variables may help a company estimate the highest possible risk (at
were used: duration of each activity and its variability, number certain probabilities) that it may be exposed to.
and distribution of transaction, and value of each transaction. While the Supply Chain Council defined VaR in e-
The main outputs were the lead time and cost of each trans- procurement as the sum of the probability of events times the
action, and the full-time equivalents (FTEs) required at each monetary impact of the events for the specific process, supplier,
working position (e.g., procurement worker and supervisor). product, or customer [66], no approach to its measurement was
The validation and verification of the obtained outputs are presented. It was also claimed that calculating VaR from his-
vital [58]. In our case, the outputs were first validated with torical data requires a large database of events and metrics, and
company employees. In addition, the results were cross-checked can be computationally intensive [66]. Since it seems that the
with data from SAP, expert opinions, and previous research importance of VaR estimation (see usage in, e.g., [62], [65],
results. If a large discrepancy is found, it may be necessary and [67]) in the supply-chain management context is increas-
to correct the developed process model (activity 3) or recheck ing, it is important to develop more sophisticated approaches to
the collected data (activity 5). The data (value of transaction, its measurement. Our approach is an initial, yet important step
work, and delay time of each activity, etc.) about each simulated in this direction.
transaction were recorded for further analyses. The collected The aggregated results are shown in Table III. The average
data enabled an analysis of both average times and costs, and lead time decreases with the increase in automation (17% for
their distribution and the application of the VaR concept. scen2; 26% for scen3), while the average costs for orders and
Finally, all results were exported to MS Excel where they the number of procurement workers dropped considerably (e.g.,
were analyzed (activity 9; results presented in the next section). from 51 in scen1 to 32 in the third scenario), with the automation
Finally, additional assumptions were made in order to acquire of approximately half the orders. This finding is in line with the
the probabilistic distribution required to calculate the VaR (ac- fact that IT plays a vital role in eliminating the need for human
tivity 10; results are presented in the next section). If required, resources to perform routine purchasing tasks [68], and there-
a further analysis (including the use of other simulation tools) fore, online procurement is significantly positively correlated
could be undertaken (activity 11). with a higher productivity growth rate [33].
The comparison of scen3 and scen4 offers interesting in-
V. RESULTS AND ANALYSIS sights. Both scens3 and 4 are upgrades of scen2. The difference
While the proposed methodology enables an estimation of is that scen3 focuses on further information support (a larger
different aspects, our analysis focuses on the main critical suc- percent of automated orders), while scen4 focuses on organiza-
cess factors of implementing e-procurement, namely the costs, tional changes (a lower number of approvals needed). As shown
lead times, and risks. The following are the chief results of the in Table III, both scenarios contribute to a further decrease in
simulations. both costs and lead times (compared to scen2); however, scen3
1) TCs—Measured by costs of employees’ work (FTE) in the contributes more to the decrease in costs. IT support has mainly
following functions: “Procurement worker,” “supervisor, decreased the costs/workload of employees, while the organiza-
and “top manager.” An estimation of total costs (in USD) tional changes in scen4 (change in approval levels) considerably
is also calculated based on the assumption that the FTE reduce the lead time due to the elimination of waiting for su-
cost is USD 100 000 for each procurement worker, USD pervisor/manager approval. It also contributed to a decrease in
150 000 for a supervisor, and USD 300 000 for a top costs (compared to scen2), but not so much as scen3.
manager. These figures are based on our interviews and in These results are in line with earlier studies, which may in-
line with recent research [59]. dicate (although not prove) that the simulation model is cor-
2) Lead (cycle) time: A lead-time reduction is important since rect and it can be repeated (with different data) in other com-
it allows the lowering of safety stock requirements and panies. Some of the estimates of costs and lead times are
improving of customer service [60]. In addition, the vari- slightly lower-–this is because we only studied a partial transi-
ability in process outcomes is connected to uncertainty tion toward e-procurement with the automatization of a cer-
and risks. Process uncertainty is likely to be reflected in tain percentage of orders, which is a more realistic case in
late deliveries and poor quality performance, so both fast practice.
and reliable deliveries are vital [61]. Both the average However, the average lead times only reveal part of the
and distributions of lead times for each scenario were story. The variability of lead times also has to be studied since
examined. it can pose even greater problems at the supply-chain level.
3) VaR: A novel application of VaR is proposed in order It is often claimed that the core goal of problem solving in
to estimate the potential risk exposure due to organiza- procurement/supply chain management is to reduce uncertain-
tional changes. VaR is defined as the expected loss aris- ties [69], [70]. Companies usually respond to time variability
ing from an adverse market movement with a specified by increasing time buffers, which is reasonable and easy to use,
probability over a period of time [62]. While VaR was yet highly inefficient [71]. The probabilistic distribution of lead
primarily intended to measure the risk of exposure in the times for all four scenarios is shown in Fig. 4, while Table IV
financial industry [63], [64], it has been not applied to a shows the likelihood that the lead time of a certain order will be
great extent to engineering systems, in general, and supply shorter than 8/14 days. For example, for scenario 1, the likeli-
chain/procurement in particular [65]. Such an approach hood that the lead time of the transaction would be eight days or

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
344 IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT, VOL. 57, NO. 2, MAY 2010

Fig. 4. Probabilistic distribution of the lead times of transactions (all four scenarios).

TABLE IV USD 100 000 (still amounting to 51.8% of the total procured
PROBABILITIES THAT A TRANSACTION WILL FINISH WITHIN A GIVEN PERIOD
OF TIME
value) and 3.22% over USD 10 000 (72.1% of the total procured
value). On the other hand, Chemicalia had several large orders.
4.0% orders (95.6% of the total value) of their orders were over
USD 100 000 and 12.0% order over USD 10 000 (99.2% of the
total value).
Simulation wise, this means that (all other things equal) the
lead times (shown in, e.g., Fig. 4 for Judril) of Chemicalia are
slightly longer and differently distributed (with a larger “long
tail” on the right) than those of Judril especially. This is due to
a longer approval process (more orders need to be approved).
Similarly, the approval costs (shown in Table III for Judril) are
less is 5% (with an 89% likelihood that it would be shorter than higher. A number of large orders also increases the VaR due to a
14 days, meaning that 84% of transactions would take between higher possible impact due to problems with a few large orders.
8 and 15 days). For, e.g., scenario 4 (where the expected lead As can be seen from the distributions, introduction of the
time is much shorter), the likelihood of filling an order within PCard and e-catalog does allow the quick processing of some
five days is 36% (and 96% for 14 days). purchase orders. It also considerably decreases the average lead
The results are important for companies that plan their in- time. However, their introduction does not considerably affect
ventory safety levels based on expected lead times. Since the the long tail—orders that last 14 days or more. These results
technological and organizational changes reduced both the lead confirm the finding that at a lower level of process maturity,
time and its variability, it is logical that the likelihood of ex- the P2P process uses some automation, but is still unpredictable
ecuting an order within the 8/14-day time period is larger for with over half of the purchases in a time-consuming process that
scen2–scen4 and the occurrence of out-of-stock situations is is largely uncontrolled [6].
lower (assuming that the safety buffers are unchanged). If, for Organizational changes, namely the empowerment of the em-
example, a company uses a 14-day time buffer, it can be assumed ployees (scen4; tripling their authorization level) achieves all of
that it will run out of stock during 11% (scen1), 9% (scen2), 9% the aforementioned. In addition to shorter average lead times,
(scen3), and 4% (scen4) of the orders. the number of transactions with extremely long lead times is re-
While the value of orders for the simulation was estimated duced. Also, the workload of managers/supervisors is reduced,
based on the data from Judril, the input data of two companies allowing them to focus more on value-added activities. The
were also analyzed (Cementy with 36 440 orders and Chemicalia average costs and the number of purchasing workers do not
with 50 612 orders in the database). The main difference was change considerably since they still perform the same tasks (but
found in the distribution of the value of orders. Cementy mainly their waiting for approvals that caused delays is considerably
had a larger number of smaller orders—only 0.31% orders over reduced).

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
TRKMAN AND MCCORMACK: ESTIMATING THE BENEFITS AND RISKS OF IMPLEMENTING E-PROCUREMENT 345

TABLE V
TOTAL VALUE OF TRANSACTIONS THAT DID NOT NEED APPROVAL LEVELS 1 AND 2

However, such changes also bring potential risks. Table V Obviously, the numerical results of this experiment cannot be
shows a considerable increase in the value of transactions with generalized to all companies since they will have different pro-
no supervisor approval. The value of transactions with no ap- ductivity profiles [74]. Savings, realized by other adopters, do
proval increases by 392% and the value of transactions without not ensure substantial cost savings for every firm [31]. However,
approval level 2 increases by 4%. This shows that the reduc- a similar approach (with different data) can be repeated for most
tion in lead time was mostly realized through the elimination of companies. The simulation model is also flexible enough to be
middle management (approval level 1). This is in line with the extended with the inclusion of other sorts of costs, if necessary.
finding that e-procurement can drastically reduce the number Further, VaR was also calculated from the results of all simu-
of middle managers needed (see, e.g., [72]) and waiting for ap- lations. First, the difference between value loss and VaR needs
proval can considerably extend the lead times when not having to be defined; the value loss is a simple arithmetic average,
a large impact on TCs [73]. However, it may pose risks of wrong meaning the expected average loss in the process (in our case,
orders emerging due to mistakes (or even fraud) by procurement due to mistakes and the fraud of employees). However, the VaR
workers. calculation demands the probabilistic distribution of potential
Therefore, an additional analysis was conducted on the as- outcomes; in our case, a normal distribution of the employees’
sumption that lower approval levels increase the chance of a loss mistakes was assumed (as outlined earlier). Other distributions
due to mistakes/incorrect orders. Assuming that the value of em- could also be used with the same approach to the VaR calcu-
ployees’ mistakes is the following percentage of each order: lation. A standard normal table for the normal distribution was
1) 3% (2% standard deviation; left truncated at 0%) if the used to calculate the threshold (from all simulated data), which
order is not authorized; constitutes the VaR at 95% (99%). This is the dollar value from
2) 1.5% (0.5% standard deviation; left truncated at 0%) if the which 95% (99%) of expected losses in different realizations of
order receives authorization from the supervisor; and the simulation are lower (this is the 95% and 99% percentiles
3) 0.5% (0.2% standard deviation; left truncated at 0%) if the from the distribution).
order receives authorization from both the supervisor and The approach to the VaR calculation, in general, is that the
the top manager. probabilistic distribution of potential losses needs to be gen-
Table VI and Fig. 5 show the simulated total costs as the sum erated first. This can be either with a similar simulation-based
of approval and wrong order costs for various approval levels. approach as found in this paper or with one of the other methods
The x scale is indexed with the current approval costs having the that consider the likelihood of various events and their impact
index 100. Approval costs (the dotted line in Fig. 5) are the cost (see, e.g., [75]); e.g., the data could also be obtained from in-
of the work of the supervisor/top managers. Value loss costs terviews with procurement workers and managers. After the
(the dashed line) are the costs arising due to wrong orders. Total creation of such a distribution, the dividing line between, e.g.,
costs (the solid line) are the sum of both. 95% of lower and 5% of higher values should be taken as a VaR
A high level (or nonexisting level) of approval obviously at 95%. Table VII shows the increase of VaR in the case of the
leads to a laissez-faire organization, where most orders are made empowerment of employees (scen4 versus scen2).
without any supervision. This obviously means low approval It should be emphasized that VaR does not measure the ex-
costs but a high value loss due to wrong orders. On the other pected losses but the probable maximum losses that a company
hand, a low approval level (a very bureaucratic organization) may accrue with a certain strategy. The number USD 2 052 932
reduces the number of wrong orders but drastically increases means that (in scen2) there is a 95% chance that the loss due
the approval costs. to wrong orders will be below this amount. In scen4, the loss
First, such an approach enables an analysis of the impact due to wrong orders will be below USD 3 740 071 (with a 95%
of changes in approval level on approval and value loss costs. probability). This example clearly illustrates the tradeoff: lower
In our case, a 10% increase in the approval level brings a 5% control costs mean a higher VaR and consequently higher risks.
decrease in approval costs and a 2% increase in value loss; a The main limitation of the VaR approach is that it is a rather
10% decrease would bring an 8% increase in approval costs and simplistic approach and provides an insight into expected losses
a 3% decrease in value loss. In such a way, the optimal approval in “normal” business conditions, but is inappropriate for ana-
level can also be found, namely where the total costs are the lyzing the impact of truly catastrophic events with a low proba-
lowest. In our example, it is reached with a 250% increase in bility [76]. Generally, organizations plan to protect themselves
the approval level (index 350). Obviously, these results should against recurrent, low-impact risks in their supply chains, but
only be used as guidance in decision making as they are only ignore high-impact, low-likelihood risks [77], [78]. In our ex-
an estimation of the tradeoff between benefits and risks (as also ample, the costs due to supplier bankruptcy or employee fraud
shown with the VaR calculation in the continuation of the paper). are not included. In order to include these risks also, the VaR

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
346 IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT, VOL. 57, NO. 2, MAY 2010

TABLE VI
TOTAL PROCUREMENT COSTS FOR DIFFERENT APPROVAL LEVELS

Fig. 5. Procurement costs (approval, depending on the value of approval level 1).

TABLE VII
VAR OF THE DIFFERENT SCENARIOS

approach should be complemented with other risk-management must include the whole P2P process that usually spans various
practices (see, e.g., [75]). departments and managerial levels. However, it was often found
that local optima do not lead to a global optimum (e.g., [79]),
and past experience has shown that many efforts have failed
VI. DISCUSSION AND CONCLUSION because they targeted processes contained only in a single de-
The procurement process is one of the most important pro- partment [17]. Therefore, our approach is holistic and includes
cesses, and its costs, reliability, and risks considerably influence the whole P2P process.
the performance or even survival of a company. Still, many com- Further, due to differences in procurement processes, activity
panies lack an approach to rigorously and quantitatively evaluate times, and the number and relations with suppliers, self-reported
their options, benefits, and risk. First of all, such an approach average results cannot be applied in process redesign and

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
TRKMAN AND MCCORMACK: ESTIMATING THE BENEFITS AND RISKS OF IMPLEMENTING E-PROCUREMENT 347

e-procurement technology implementation in a specific com- conceived notion of the benefits effects the adoption of e-
pany. A company should carefully analyze its own business procurement) and the implementation phase (to monitor the
processes and procurement transactions, and use suitable crite- project and the achievement of expected benefits). The results
ria (e.g., costs, risks, lead time, and percentage of transactions are explained in the terminology commonly accepted by finan-
beyond a certain threshold). Both the criteria and the simu- cial managers (e.g., costs in dollars, FTE, VaR). It confirms the
lation/process model can be properly modified without much possibility to establish a conceptual link between financial con-
trouble. cepts and process management [62]. Further, the results from
In the event, another studied company was to have a slightly process modeling and simulation can serve as inputs for the use
different procurement process and the process/simulation model of activity-based costing [81].
should also be properly adapted. The process model (Fig. 3) is The research has various limitations. First, the procurement
developed broadly enough to serve as a reusable framework process scenarios were deliberately simplified in order to allow
that can be applied to other similar procurement processes with a focus on the main problems and minimize interaction effects
the application of the approach shown in Fig. 2. However, the that may mask the results for that scenario. They do not include
approach is unsuitable for companies that rely on automated the suppliers’ activities, their connection to the buyer, and the
ordering with the use of just in time, vendor management in- potential delays or disruptions due to problems in the suppliers’
ventory, or similar concepts. Such an approach is often found internal business processes or transport routes.
in, e.g., the automotive industry. The preparedness of suppliers to implement joint e-
Therefore, the paper’s main contribution is its presentation procurement solutions should also be studied. The calcula-
of a methodological approach to the measurement of risks and tion of VaR only included the probabilistic distribution of the
benefits of implementing e-procurement with an analysis of percentage of employees’ mistakes in the calculation. In or-
different scenarios. The results show that the same decrease der to provide a comprehensive estimation of risks, external
in costs and lead times does not necessarily happen simulta- risks should also be included (e.g., logistics problems, supplier
neously, and organizational changes/process improvement can nonperformance).
often bring even greater savings than implementation of a simple Also, the research only focused on an estimation of process
technology. The results can be used as benchmark/key perfor- costs but not on possible changes in purchase prices due to
mance indicators for monitoring an e-procurement implemen- implementing e-procurement. Depending on the type of product,
tation project. the savings can range between 7% and 17% [29], [82]. Our
A novel approach to the optimization of an organizational approach also did not include an estimate of the investment
strategy (e.g., approval level) is also presented and enables the costs (both capital investment in technology and the necessary
finding of the optimal point in a tradeoff between the costs of effort and costs of changes in the organizational structure and
approval procedures and the potential costs due to mistakes or employees’ roles) and the costs of operating and maintaining
fraud in the case the procedure is simplified. Business process the e-procurement system.
models and simulations enable an ex ante analysis of the impact While our study did not use underlying theory, the proposed
of such changes, before any changes are actually made. approach enables further investigations within either TCs or in-
The presented approach is most suitable for a company on formation processing (IP) view theory. The TC theory is namely
either level 2 (defined) or level 3 (linked) of a five-level pro- the most frequently applied theory in e-procurement studies [2].
cess maturity model—most companies are currently on these Specifically, the ability to estimate both TC and the risks of e-
levels [80]. Such companies have both the required data and procurement enables the utilization of risk-augmented TC the-
the developed process maps needed for the preparation of the ory [35] to analyze the effects of organizational changes (e.g.,
simulation model. Companies on level 1 (ad hoc) usually do the empowerment of employees and reduction of middle man-
not possess detailed enough data about each activity/process, agement) on both TCs and risks.
process maps, or metrics at the process level. Companies on Alternatively, the IP view of interorganizational coordination
level 4 or 5 take their cooperation with suppliers to the process [83] could be used to assess which process configuration is most
level and the model in Fig. 1 should be expanded to integrate suitable for collecting and processing information. IP needs are
suppliers’ activities (see, e.g., [11]). assessed based on various characteristics of the product and
Since risks in e-procurement are also important, the novel procurement environment, while IP capabilities are assessed by
application of the VaR concept in procurement research can the level of IT support for various activities in the procurement
enable the monitoring of these risks and (if necessary) the jus- life cycle [84]. Simulation results can be used for that purpose:
tification of the acceptance of mitigatory actions. The paper variability of lead times/costs can be a proxy for uncertainty,
namely presents one of the first approaches to measure the VaR while different scenarios analyze the effects of different levels
of the procurement process in monetary terms. The use of VaR of IT support.
can also improve benchmarking between processes and compa-
nies, and contribute to the development of a common language
for studying procurement risks.
The approach has several managerial implications. It can ACKNOWLEDGMENT
serve both in the project preparation (to estimate the poten- The authors would like to thank the Department Editor Dr.
tial benefits and justify the investment; additionally, the pre- Ravichandran and the two anonymous referees for their several

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
348 IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT, VOL. 57, NO. 2, MAY 2010

useful comments that improved both the content and presenta- [25] L. de Boer, J. Harink, and G. Heijboer, “A conceptual model for assess-
tion of the paper. ing the impact of electronic procurement,” Eur. J. Purchasing Supply
Manage., vol. 8, pp. 25–33, 2002.
[26] J. Gansler, W. Lucyshyn, and K. Ross, Digitally Integrating the Gov-
ernment Supply Chain: E-Procurement, E-Finance, and E-Logistics.
Washingon, DC: IBM Endowment for the Business of Government, 2003.
REFERENCES [27] T. Puschmann and R. Alt, “Successful use of e-procurement in supply
chains,” Supply Chain Manage.—Int. J., vol. 10, pp. 122–133, 2005.
[1] J. Hartley, M. Lane, and Y. Hong, “An exploration of the adoption of
[28] W. D. Presutti, “Supply management and e-procurement: Creating value
E-auctions in supply management,” IEEE Trans. Eng. Manag., vol. 51,
added in the supply chain,” Ind. Marketing Manage., vol. 32, pp. 219–226,
no. 2, pp. 153–161, May 2004.
2003.
[2] T. Schoenherr and R. Tummala, “Electronic procurement: A structured
[29] T. Minahan, The E-procurement Benchmark Report. Boston, MA: Ab-
literature review and directions for future research,” Int. J. Procurement
erdeen Group, 2004.
Manage., vol. 1, pp. 8–37, 2007.
[30] L. Giunipero and R. Eltantawy, “Securing the upstream supply chain:
[3] F. Wu, G. A. Zsidisin, and A. Ross, “Antecedents and outcomes of E-
A risk management approach,” Int. J. Phys. Distrib. Logist. Manage.,
procurement adoption: An integrative model,” IEEE Trans. Eng. Manag.,
vol. 34, pp. 698–713, 2004.
vol. 54, no. 3, pp. 576–587, Aug. 2007.
[31] D. Hur, V. A. Mabert, and J. L. Hartley, “Getting the most out of reverse
[4] Y.-W. Yu, H.-C. Yu, H. Itoga, and T.-R. Lin, “Decision-making factors for
e-auction investment,” Omega, vol. 35, pp. 403–416, 2007.
effective industrial e-procurement,” Technol. Soc., vol. 30, pp. 163–169,
[32] Aberdeen Group, Accounts Payable Supplier Enablement: The Best-in-
2008.
Class Advantage. Boston, MA: Harte-Hanks, 2008.
[5] T. M. Rajkumar, “E-Procurement: Business and technical issues,” Inf.
[33] M. Falk, “ICT-linked firm reorganisation and productivity gains,” Tech-
Syst. Manage., vol. 18, pp. 1–9, 2001.
novation, vol. 25, pp. 1229–1250, 2005.
[6] R. Handfield, K. McCormack, and W. Steininger, “Best practices in pro-
[34] T. Osmonbekov, D. Bello, and D. Gilliland, “Adoption of electronic com-
cure to pay,” NC State University, Raleigh, 2005.
merce tools in business procurement: Enhanced buying center structure
[7] A. Davila, M. Gupta, and R. Palmer, “Moving procurement systems to
and processes,” J. Bus. Ind. Marketing, vol. 17, pp. 151–166, 2002.
the internet: The adoption and use of e-procurement technology models,”
[35] R. Kauffman and H. Mohtadi, “Proprietary and open systems adoption
Eur. Manage. J., vol. 21, pp. 11–23, 2003.
in e-procurement: A risk-augmented transaction cost perspective,” J.
[8] P. Davamanirajan, R. Kauffman, C. Kriebel, and T. Mukhopadhyayd,
Manage. Inf. Syst., vol. 21, pp. 137–166, 2004.
“Systems design, process performance, and economic outcomes in inter-
[36] H. Min and W. P. Galle, “E-purchasing: Profiles of adopters and non-
national banking,” J. Manage., vol. 23, pp. 65–90, 2006.
adopters,” Ind. Marketing Manage., vol. 32, pp. 227–233, 2003.
[9] A. Ordanini and G. Rubera, “Strategic capabilities and internet resources
[37] C. K. Riemenschneider, D. A. Harrison, and P. P. Mykytyn, “Understand-
in procurement: A resource-based view of B-to-B buying process,” Int.
ing it adoption decisions in small business: Integrating current theories,”
J. Oper. Product. Manage., vol. 28, pp. 27–52, 2008.
Inf. Manage., vol. 40, pp. 269–285, 2003.
[10] A. Soares-Aguiar and A. Palma-dos-Reis, “Why do firms adopt e-
[38] R. Agarwal and J. Prasad, “A field study of the adoption of software
procurement systems? Using logistic regression to empirically test a con-
process innovations byinformation systems professionals,” IEEE Trans.
ceptual model,” IEEE Trans. Eng. Manag., vol. 55, no. 1, pp. 120–133,
Eng. Manage., vol. 47, no. 3, pp. 295–308, Aug. 2000.
Feb. 2008.
[39] K. A. Saeed and S. Abdinnour-Helm, “Examining the effects of infor-
[11] P. Trkman, M. I. Štemberger, J. Jaklič, and A. Groznik, “Process approach
mation system characteristics and perceived usefulness on post adop-
to supply chain integration,” Supply Chain Manage.—Int. J., vol. 12,
tion usage of information systems,” Inf. Manage., vol. 45, pp. 376–386,
pp. 116–128, 2007.
2008.
[12] G. Hunter, M. Bunn, and W. Perreault, “Interrelationships among key
[40] C. Lin and G. Pervan, “The practice of IS/IT benefits management in large
aspects of the organizational procurement process,” Int. J. Res. Marketing,
Australian organizations,” Inf. Manage., vol. 41, pp. 13–24, 2003.
vol. 23, pp. 155–170, 2006.
[41] S. S. Rao, D. Truong, S. Senecal, and T. T. Le, “How buyers’ ex-
[13] T. Ravichandran, S. Pant, and D. Chatterjee, “Impact of industry structure
pected benefits, perceived risks, and e-business readiness influence their
and product characteristics on the structure of B2B vertical hubs,” IEEE
e-marketplace usage,” Ind. Marketing Manage., vol. 36, pp. 1035–1045,
Trans. Eng. Manag., vol. 54, no. 3, pp. 506–521, Aug. 2007.
2007.
[14] L. Ellram and G. A. Zsidisin, “Factors that drive purchasing and supply
[42] A. N. Mishra, P. Konana, and A. Barua, “Antecedents and consequences
management’s use of information technology,” IEEE Trans. Eng. Manag.,
of internet use in procurement: An empirical investigation of U.S. manu-
vol. 49, no. 3, pp. 269–281, Aug. 2002.
facturing firms,” Inf. Syst. Res., vol. 18, pp. 103–120, 2007.
[15] J. Y. Bakos, “Reducing buyer search costs: implications for electronic
[43] H. Shin, D. A. Collier, and D. D. Wilson, “Supply management orientation
marketplaces,” Manage. Sci., vol. 43, pp. 1676–1692, 1997.
and supplier/buyer performance,” J. Oper. Manage., vol. 18, pp. 317–333,
[16] S. Dunlap, “The last unautomated frontier: How technology is stream-
2000.
lining the invoice-to-cash process,” AFP Exch., vol. 2005, pp. 14–17,
[44] D. Graham, S. Smith, and P. Dunlop, “Lognormal distribution provides an
2005.
optimum representation of the concrete delivery and placement process,”
[17] A. Spanyi, More for Less: The Power of Process Management. Tampa,
J. Constr. Eng. Manage., vol. 131, pp. 230–238, 2005.
FL: Meghan-Kiffer Press, 2007.
[45] J. Lee, K. Siau, and S. Hong, “Enterprise integration with ERP and EAI,”
[18] P. E. D. Love, Z. Irani, C. Standing, C. Lin, and J. M. Burn, “The enigma
Commun. ACM, vol. 46, pp. 54–60, 2003.
of evaluation: Benefits, costs and risks of IT in Australian small-medium-
[46] J. P. Baron, M. J. Shaw, and A. D. Bailley, Jr., “Web-based E-catalog
sized enterprises,” Inf. Manage., vol. 42, pp. 947–964, 2005.
Systems in B2B Procurement,” Commun. ACM, vol. 43, pp. 93–100,
[19] C. Subramaniam and M. Shaw, “The effects of process characteristics on
2000.
the value of B2B E-procurement,” Inf. Technol. Manage., vol. 5, pp. 161–
[47] E. Boulianne, “The impact of procurement card usage on cost reduction,
180, 2004.
management control, and the managerial audit function,” Managerial
[20] A. Gunasekaran and E. W. T. Ngai, “Adoption of e-procurement in Hong
Aud. J., vol. 20, pp. 592–604, 2005.
Kong: An empirical research,” Int. J. Product. Econ., vol. 113, pp. 159–
[48] R. Handfield, Supply Market Intelligence: A Managerial Handbook for
175, 2008.
Building Sourcing Strategies. Boca Raton, FL: Auerbach, 2006.
[21] C. Tanner, R. Wolfle, P. Schubert, and M. Quade, “Current trends and
[49] M. Caridi, S. Cavalieri, G. Diazzi, and C. Pirovano, “Assessing the impact
challenges in electronic procurement: An empirical study,” Electron.
of e-procurement strategies through the use of business process modelling
Markets, vol. 18, pp. 6–18, 2008.
and simulation techniques,” Prod. Plann. Control, vol. 15, pp. 647–661,
[22] K. Reeves, “Supply chain governance: A case of cross dock management
2004.
in the automotive industry,” IEEE Trans. Eng. Manag., vol. 54, no. 3,
[50] A. Greasley, “Using process mapping and business process simulation to
pp. 455–467, Aug. 2007.
support a process-based approach to change in a public sector organisa-
[23] K. McCormack, Business Process Maturity: Theory and Application.
tion,” Technovation, vol. 26, pp. 95–103, Jan. 2006.
North Charleston, SC: Booksurge LLC, 2007.
[51] E. Mohebbi, F. Choobineh, and A. Pattanayak, “Capacity-driven vs.
[24] C. Ash and J. Burn, “Assessing the benefits from e-business transformation
demand-driven material procurement systems,” Int. J. Prod. Econ.,
through effective enterprise management,” Eur. J. Inf. Syst., vol. 12,
vol. 107, pp. 451–466, 2007.
pp. 297–308, 2003.

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
TRKMAN AND MCCORMACK: ESTIMATING THE BENEFITS AND RISKS OF IMPLEMENTING E-PROCUREMENT 349

[52] A. Roeder and B. Tibken, “A methodology for modeling inter-company [79] T. Kobayashi, M. Tamaki, and N. Komoda, “Business process integration
supply chains and for evaluating a method of integrated product and as a solution to the implementation of supply chain management systems,”
process documentation,” Eur. J. Oper. Res., vol. 169, pp. 1010–1029, Inf. Manage., vol. 40, pp. 769–780, 2003.
2006. [80] K. McCormack and W. Johnson, Business Process Orientation: Gaining
[53] K. Vergidis, A. Tiwari, and B. Majeed, “Business process analysis and the E-Business Competitive Advantage. Delray Beach, FL: St Lucie
optimization: Beyond reengineering,” IEEE Trans. Syst., Man, Cybern. Press, 2001.
C, Appl. Rev., vol. 38, no. 1, pp. 69–82, Jan. 2008. [81] M. Zhang and M. Tseng, “A product and process modeling based approach
[54] J. Berlak and V. Weber, “How to make e-Procurement viable for SME to study cost implications of product variety in mass customization,” IEEE
suppliers,” Prod. Plann. Control, vol. 15, pp. 671–677, 2004. Trans. Eng. Manag., vol. 54, no. 1, pp. 130–144, Feb. 2007.
[55] I. Davies, P. Green, M. Rosemann, M. Indulska, and S. Gallo, “How do [82] E. Bartezzaghi and S. Ronchi, “E-sourcing in a buyer-operator-seller per-
practitioners use conceptual modeling in practice?,” Data Knowl. Eng., spective: Benefits and criticalities,” Prod. Plann. Control, vol. 16, pp. 405–
vol. 58, pp. 358–380, 2006. 412, 2005.
[56] A. Groznik, A. Kovačič, and P. Trkman, “The role of business renovation [83] M. Bensaou and N. Venkatraman, “Inter-organizational relationships and
and informatization in e-government,” J. Comput., vol. 49, pp. 80–88, information technology: A conceptual synthesis and a research frame-
2008. work,” Eur. J. Inf. Syst., vol. 5, pp. 84–91, 1996.
[57] J. Erjavec, M. Gradisar, and P. Trkman, “Renovation of the cutting stock [84] G. Premkumar, K. Ramamurthy, and C. S. Saunders, “Information pro-
process,” Int. J. Prod. Res., vol. 47, pp. 3979–3996, 2008. cessing view of organizations: An Exploratory examination of fit in the
[58] F. Persson and J. Olhager, “Performance simulation of supply chain de- context of interorganizational relationships,” J. Manage., vol. 22, pp. 257–
signs,” Int. J. Prod. Econ., vol. 77, pp. 231–245, 2002. 294, 2005.
[59] S. Avery, “Purchasing 2007 salary survey: Purchasing salaries continue
their climb,” Purchasing, 2007.
[60] C. Chandra and J. Grabis, “Inventory management with variable lead-time
dependent procurement cost,” Omega, vol. 36, pp. 877–887, 2008.
[61] S. Vachon and R. Klassen, “An exploratory investigation of the effects
of supply chain complexity on delivery performance,” IEEE Trans. Eng.
Manag., vol. 49, no. 3, pp. 218–230, Aug. 2002. Peter Trkman received the M.Sc. and Ph.D. degrees
[62] C. S. Tapiero, “Value at risk and inventory control,” Eur. J. Oper. Res., in management information systems from the Uni-
vol. 163, pp. 769–775, 2005. versity of Ljubljana, Ljubljana, Slovenia.
[63] S. Basak and A. Shapiro, “Value-at-risk-based risk management: Optimal He is currently an Assistant Professor in the
policies and asset prices,” Rev. Finan. Stud., vol. 14, pp. 371–405, 2001. Faculty of Economics, University of Ljubljana. His
[64] K. Dowd, Beyond Value at Risk: The New Science of Risk Management research interests include technology adoption, e-
(Wiley Series in Financial Engineering). New York: Wiley, 1998. government, and various aspects of the supply chain,
[65] D. Simchi-Levi, “Managing uncertainty and risk in the supply chain using business process, and operations management. He
financial engineering instrument,” presented at the MIT Eng. Syst. Div. has participated in several national and international
2004 Symp., Cambridge, MA. projects (both research and consulting). He is the au-
[66] D. Kent, “Supply chain risk measurement,” Supply-Chain Council Risk thor or coauthor of more than 60 papers/book chap-
Management Team, Shangai, China, 2007. ters, including papers in the Computers and Operations Research, the European
[67] B. Tomlin, “On the value of mitigation and contingency strategies for Journal of Operational Research, the Government Information Quarterly, the
managing supply chain disruption risks,” Manage. Sci., vol. 52, pp. 639– International Journal of Information Management, the International Journal of
657, 2006. Production Economics, the International Journal of Production Research, the
[68] A. Carr and L. Smeltzer, “The relationship between information technol- Journal of Computer Information Systems, the Online Information Review, the
ogy use and buyer—Supplier relationships: An exploratory analysis of Supply Chain Management—An International Journal, the Technology Fore-
the buying firm’s perspective,” IEEE Trans. Eng. Manag., vol. 49, no. 3, casting and Social Change, and the Telecommunications Policy.
pp. 293–304, Aug. 2002.
[69] G. Reiner and M. Trcka, “Customized supply chain design: Problems and
alternatives for a production company in the food industry. A simulation
based analysis,” Int. J. Prod. Econ., vol. 89, pp. 217–229, 2004.
[70] B. Yang, N. D. Burns, and C. J. Backhouse, “Management of uncertainty
through postponement,” Int. J. Prod. Res., vol. 42, pp. 1049–1064, 2004.
[71] K. T. Yeo and J. H. Ning, “Managing uncertainty in major equipment
procurement in engineering projects,” Eur. J. Oper. Res., vol. 171, pp. 123– Kevin McCormack received the degrees in Chem-
134, 2006. istry, Engineering, MBA, and DBA.
[72] G. Hamel, Das revolutionäre Unternehmen. (Wer Regeln bricht: Gewinnt). He is currently the President of the DRK Research,
Muenchen, Germany: Econ, 2000. Raleigh, NC. He is also associated with various Uni-
[73] C. J. Mayer and C. T. Somerville, “Land use regulation and new construc- versities such as Babson College, NC State, and the
tion,” Reg. Sci. Urban Econ., vol. 30, pp. 639–662, 2000. University of Oklahoma. He is the Master Instructor
[74] J. Harvey, L. Lefebvre, and E. Lefebvre, “Exploring the relationship be- for Supply Chain Risk with the Supply Chain Coun-
tween productivity problems and technology adoption in small manufac- cil. He has more than 30 years of business leadership,
turing firms,” IEEE Trans. Eng. Manag., vol. 39, no. 4, pp. 352–358, Nov. engineering, teaching, research, and consulting expe-
1992. rience in the areas of information technology, oper-
[75] Hackett Group, “Dow Chemical Company: Supply Risk Management ations management, and supply chain management.
Process Is Key to Improving Safety and Security,” in Proc. Third Ann. He has also developed and delivered courses in information technology, HR,
Europ. Best-Pract. Conf., “Leveraging Synergies: Myth or Reality,” May operations management, and supply chain management at the graduate and un-
10–11, 2007. dergraduate level both in U.S., China, and Europe. He is the author or coauthor
[76] D. Shi, “A review of enterprise supply chain risk management,” J. Syst. of five books and more than 100 articles in the Quality Progress, the Business
Sci. Syst. Eng., vol. 13, pp. 219–244, 2004. Process Management Journal, the Supply Chain Management, Benchmarking:
[77] S. Chopra and M. S. Sodhi, “Managing risk to avoid supply-chain break- A International Journal, and the Supply Chain Management Review. He is also
down,” Sloan Manage. Rev., vol. 46, pp. 53–61, 2004. a Judge for the Manufacturer of the Year Award for the state of Alabama, home
[78] M. Faisal, D. K. Banwet, and R. Shankar, “Mapping supply chains on risk of various international manufacturers’ locations (Honda, Mercedes, Lockheed,
and customer sensitivity dimensions,” Ind. Manage. Data Syst., vol. 106, BASF, Nucor, U.S. Steel, and Siemens Automotive) as well as dozens of defense
pp. 878–895, 2006. and automotive suppliers.

Authorized licensed use limited to: North Carolina State University. Downloaded on April 30,2010 at 18:23:54 UTC from IEEE Xplore. Restrictions apply.
View publication stats