|
Abstract
In this paper, we present a computer-based simulation exercise designed to help students understand the impact of an enterprise system on business performance in a service supply chain. The particular service supply chain simulated in the exercise is a wireless telecommunications firm. In this exercise, students perform simulations to experience managing the supply chain of the telecommunications firm with and without an enterprise system. The simulator tracks their business performance. Then the results are used as the basis of discussion in a subsequent debriefing session. We describe the educational goals of the simulation exercise and how the exercise can be structured in order to achieve these goals. The latter is illustrated by the use of the simulation exercise in a master's level supply chain management course in the Red McCombs School of Business at the University of Texas at Austin. The simulator includes realistic details. In fact, it is based on the extensive consulting experiences of the first author with two North American telecommunications firms. We describe the simulator in detail under the various scenarios, explain how it was validated, and provide the simulator equations in system dynamics format in Appendix B.
Most large organizations have implemented or are considering implementing some form of an enterprise-wide information system (ES). Proponents of these ESs suggest they provide many benefits over the legacy systems they replace, such as better systems integration, standardization of data and processes, end user visibility across the business enterprise, and improved decision support functionality (Davenport, 2000; Ettlie, 2000; Gattiker and Goodhue, 2000; Mabert et al., 2000; McAfee, 2000). ESs cost millions of dollars and take many months to implement fully (O'Leary, 2000). Thus, when evaluating different ESs, decision makers need to understand the potential benefits and costs. In practice, quantifying the impact of an ES (both costs and benefits) may be difficult due to lack of good benchmarking data and the challenge of isolating the impact of an ES from other potentially confounding factors.
This paper describes a computer-based simulation exercise designed to teach students about the potential impact of a successfully implemented enterprise system on business performance in a service supply chain. The system dynamics simulator allows students to experience the value added by running the supply chain of a wireless telecommunications firm under two different scenarios: with and without an ES. The simulator tracks the performance of decision-makers as they manage the supply chain of the simulated telecommunication firm under one of the aforementioned scenarios. Thus, this exercise isolates the impact of an ES implementation. The results from these exercises are compared to determine whether or not the proposed ES benefits influence the bottom line under the use of a typical manager. It is not always clear if an ES will help managers run dynamically complex systems (Sharda et al., 1988). Therefore, besides providing a hands-on management training experience, this exercise also provides an opportunity for a rich discussion of the issues with which companies, the academic literature, and the popular press are wrestling. A paper containing a preliminary proposal for this simulation exercise was presented at the 2000 Winter Simulation Conference (Ritchie-Dunham et al., 2000).
We have designed the simulation exercise to be a three to six-hour self-contained module that can be used as part of a longer course in supply chain management or information management, or as a stand-alone executive education short course. In Section 5, we demonstrate its use over two 75-minute sessions in a master's level supply chain management class in the Red McCombs School of Business at the University of Texas at Austin. The students performed the simulation in the first session. The second session was spent discussing the simulation results and how these results related to actual company experiences and to those studied in the academic literature.
We selected a service sector firm, in particular, a wireless telecommunications firm, as the basis for this simulation exercise for four reasons. First, our students and corporate partners have requested more service sector related exercises be added to our curriculum (Anderson and Morrice, 2000) because the majority of our graduates find employment in the service sector. Second, games and other exercises are more common for physical goods supply chains (Littlefield Technologies, Bates, 2002; e.g. The MIT Beer Distribution Game, Senge, 1990). Third, wireless telecommunications is a business with which the students can easily identify. They find the application area interesting making it easier to motivate the exercise. While there is the potential downside that this specific example might provide a narrow experience for the students, we do not believe this to be the case because the simulator contains elements (both financial and operational) common to firms in many other industries, such as those involved in other aspects of information technology services, consulting, retail services, third party logistics, and call centers.
The final reason for selecting a wireless telecommunications firm as the case example is that the first author has extensive consulting experience in the telecommunications industry. In fact, the simulator is based on system dynamics simulation models developed and used for strategic policy analysis at two North American telecommunications firms. It is well validated, comprehensive, and easy to understand (see Appendix B for simulator equations).
Throughout the paper we often use the term ES synonymously with Enterprise Resource Planning (ERP) systems especially when citing some of the more recent academic literature. We have consciously decided to use ES because we think it is more general than a specific business term used in current practice (namely, ERP). Additionally, we believe that the significance and purpose of our simulation exercise transcends the current issue of implementing Enterprise Resource Planning (ERP) systems. The business value of information technology to the enterprise has been a topic of interest in the literature for the last 15 - 20 years with more recent literature focusing on the value of specific applications such as ERP systems (Hitt et al. 2002). Based on the results of the study by Akkermans et al. (2003), this topic is likely to be of continuing interest to business executives and academics for some time into the future especially in specific application areas such as supply chain management. Therefore, the overall worth of this exercise is to help students understand the impact of the aforementioned principles of integration, standardization, visibility, and improved decision support on business performance in service supply chains.
It is important to note that the simulation exercise does not include all the details and complexities involved in implementing an ES. These are accounted for in the simulator by ES implementation costs. This was done to simplify the exercise and narrow the focus down to exploring the potential benefits of a successful ES implementation. In other words, the exercise focuses on before and after the ES implementation but not the transition between these two states. It is our experience that such focusing is necessary in order to mitigate the risk of making the exercise too complex to yield any meaningful results. To provide a proper context for the students and balance the discussion, we emphasize to them that they should view the results of this exercise as the potential for ES implementation at its best. Then we discuss case studies and other literature illustrating ES implementations that have been costly and challenging (Griffith et al., 1999; Hong and Kim, 2002; Songini, 2004). The latter serves as a launch pad for a discussion on what factors can lead to a successful ES implementation.
The rest of the paper is organized in the following manner. Section 2 details the nature of operations of a wireless telecommunications firm and how the system dynamics simulator captures that nature. Section 3 presents the enterprise systems. Section 4 outlines the validation of the simulator. Section 5 provides an example of the simulation exercise conducted with master's students at the University of Texas at Austin. Section 6 contains a concluding discussion.
The simulator presented here was abstracted from a more complex system dynamics simulator that was developed by one of the authors and used by two national telecommunication firms for strategic decision making. The original simulator was simplified to make it easier for subjects to control during the experiment. Both the original and simplified simulators were developed with experts in wireless telecommunications and supply chains (see Appendix A). The simulator parameters were then scaled to protect the confidentiality of the telecommunications firms. This scaling did not affect the dynamics of the simulator.
We briefly describe how wireless telecommunications work and how a telecommunications firm provides wireless services. The principles of wireless telecommunications are straightforward. A person uses her cell phone, which acts as a radio transceiver, to connect to a nearby base station (see Figure 1). To ensure constant and continuous cell phone coverage while the user is moving, the base stations form a network of overlapping cells. These base stations manage, send, and receive signals from the cell phones in its geographic area to a mobile telecommunications switching office (MTSO). The MTSO places calls from land-based telephones to wireless customers, switches calls between base stations as the cell phone travels across cell boundaries, and authenticates wireless customers before they make calls. Additionally, the MTSO connects the network to public networks such as Southwestern Bell and other cellular networks, which connect with land-based phones. A cellular network is constructed in stages by adding more base stations to increase network coverage and quality.
Figure 1. Overview of How Wireless Telecommunications Work
Figure 2 summarizes how a telecommunications firm provides wireless services, as modeled in the simulator. Starting in the lower-left corner, the firm makes Investment Decisions that allocate financial resources to: (1) Infrastructure (base stations), (2) Human Resources (training and hiring people), and (3) the Service Supporting Information Technology (to support call center services). These three resources influence different dimensions of Customer Satisfaction, which influences the number of customers in the Customer Base and impacts the organization's Financials. The Financials influence further investment decisions and the cycle repeats itself.
Figure 2. Overview of How a Telecommunications Firm Is Setup to Provide Wireless Services
The simulator incorporates external and internal components of the firm's supply chain, as denoted by boxes in Figure 2. External components include the suppliers (within Infrastructure) and the Customer Base. Internal components include Human Resources, Infrastructure, Service Supporting Information Technology, and Financials. This section briefly explains the model logic. Appendix B provides a full description of the simulator's equations, ordered by the boxed components in Figure 2.
In the Infrastructure component of the supply chain, as expressed in the simulator (see Figure 3), the firm places orders for base stations with the suppliers. The monthly ordering capacity is based on the annual budget for base stations and their average cost. Since the firm competes with other firms also installing base stations, the suppliers' daily building capacity available to the firm is a function of the supplier's annual building capacity and the firm's market share. This supplier building capacity is a function of its initial building capacity, its growth due to changes in our firm's demand, and its growth due to overall demand, the latter being a function of growth in the population demanding wireless services and market share. The supplier's growth in building capacity is moderated by its growth response time—how long it takes to shift its capacity once it sees a shift in demand. The building of base stations converts them from Capacity in Process to Base Stations, which are in service. The base station turn around is a ratio of the capacity in process and the rate of building base stations with a lower bound constraint on minimum construction time. After an average life, the Base Stations are retired. The number of Base Stations and their daily capacity determines the Network Daily Capacity. Capacity Utilization is then a function of this Daily Capacity and the Total Daily Usage, which is a function of the number of Customers and their average usage. Capacity Utilization influences Network Quality, which influences the Customers' Perceived Call Quality (see Customer Satisfaction component of Figure 3). The other component of perceived call quality is Network Coverage, which is a ratio of the number of active Base Stations and the number of base stations required to cover 100% of the population.
On the other end of the supply chain, in the Customer Satisfaction component, the firm provides services to customers for a fee. Satisfaction is measured as a utility function of the price paid for a perceived level of call quality and customer service, relative to both the competitive offering in the marketplace and the customer's disposable income dedicated to wireless telecommunication services. Each of the factors that determine customer satisfaction is determined endogenously, as explained in the following section. The firm's retention rate is the ratio of customer satisfaction to the competiton's customer satisfaction. In the Customer Base component, the rate at which customers are gained from or lost to the competition is a function of the retention rate and the customer's delay in acting i.e., how long it takes them to respond to changes in relative customer satisfaction based on experiences and contracts (see Figure 3). The firm gains new customers as a product of the number of new users added to the population in each time period and the firm's new user market share. The population of wireless customers is assumed to grow at a constant rate over the seven-year simulation period, with the customer's average disposable income for wireless services held constant.
Figure 3. Stock-flow Model of External Components
The Human Resources component describes the dynamics of human resource skills as employees are hired or fired, and as they are trained or their skills are made obsolete by changing technologies (see Figure 4). Employees have either relevant skills or obsolete skills. Relevant Skilled People are hired into the firm and may leave the firm by attrition or by downsizing. They also convert to Obsolete Skilled People when their skills become obsolete, which is a function of the Industry Technology Change Rate. As Obsolete Skilled People, they leave by attrition or by downsizing, or they are trained to become Relevant Skilled People. The rate of people being trained is influenced by training effectiveness, which is a function of the training budget per employee with obsolete skills, as benchmarked against industry standards. Training effectiveness is also influenced by HR's ability to effectively schedule the right training when needed by the right people. The HR Service Index, a determinant of perceived Customer Service (in Figure 3), is a function of the actual customer to employee ratio and the benchmark service level that can be provided at different ratios of customers to employees.
The Service Supporting Information Technology component depicts the dynamics of the information systems that support the call centers. The firm invests its budget in new SSIT, which is retired a specified time later. The effective IT$ per employee is a ratio of the amount of SSIT available to the number of call center employees, which is moderated by the information processing quality. As information processing quality increases, the effective IT$ per employee increases for the same investment in SSIT. The effective IT$ per employee is benchmarked against industry standards to determine the degree to which the employees have sufficient IT to facilitate their ability to see the information necessary to address customer needs and act on them. Thus, the IT Facilitation Index compares the effective IT budget per call center employee to a benchmark level. The IT Facilitation Index along with the HR Service Index influence Customer Satisfaction through perceived customer service (see Figure 3).
The Financials component (see Figure 2) calculates the financial statements based on the status and flow of resources in the firm. This component is not shown in a diagram (see Appendix B for formulations).
Figure 4. Stock-flow Model of Internal Components
This section explores how enterprise systems influence the management of a complex system, like the wireless telecommunications firm overviewed above. As organizations determined that information technology might facilitate their business, they created discrete information systems for that specific purpose. As the number of these systems multiplied and the size of organizations grew, the complexity of these loosely interfaced systems increased. To link these systems more tightly, software packages began to offer integrated systems, such as material requirements planning, which provided multiple functional applications with a common database. (For a system dynamics study of these early systems, see Morecroft, 1983). These systems eventually incorporated a full suite of supply chain applications to the manufacturing systems. These systems, which began as back office support, now support the front office across the supply chain (Davenport, 2000). These enterprise systems are very complex, in their integration of many functional applications and best practices, as well as their need to evolve with the organization over time (Markus and Tanis, 2000). This complexity of interconnected functional applications and their associated organizational issues makes successful implementation of these systems very difficult. Organizations invest in these complex systems because of the promise of reduced cycle time, faster information transactions, improved financial management, ability to offer electronic commerce, and making tacit process knowledge explicit (Davenport, 2000; Hitt et al., 2002; Markus and Tanis, 2000). It is assumed that these benefits will convert eventually into higher economic return (Andersen, 1999).
To allow students to experience the value added by applying a successfully implemented ES to a firm, we create two versions of the simulator described in Section 2 to represent the following scenarios: the firm with an ES and the firm without an ES. We consider the simulator of the firm without an ES first because virtually all firms start here before installing an ES.
In the simulator of the firm without an ES, information is maintained in independent (or fragmented) information systems corresponding to the firm resources. Strategic-level performance indicators focus on financials and operations. Financial indicators include economic value added (EVA) and the components of EVA, as well as capital invested, debt to equity ratio, and current budgetary allocations. Operational indicators include average monthly orders, inventory, number of employees, market share, and number of customer complaints. To emulate independent systems, the simulation lags the information by a three-month delay and provides it in financial and operational terms.
The simulator of the firm with an ES provides additional benefits, listed by component in Table 1. In the Financials component, the administrative overhead is lower with an ES because much of the data processing is automated (Davenport, 2000). In the Human Resources component, the ES tracks the relevant skills of thousands of employees, more efficiently identifying those employees requiring training and scheduling their availability for specific courses. Additionally, by providing employees with relevant, integrated information from across the firm, the ES reduces the skills required to find and process the required information.
In the Infrastructure component, the ES's capital asset management software monitors system-wide utilization of the base stations, spreading usage more evenly over the network and planning infrastructure preventive maintenance more optimally. This reduces the overhead for service technicians waiting to respond to failed systems. In the Service Supporting Information Technology (SSIT) component, the ES integrates information across many areas, giving the employees the right information they need. The initial cost of the ES is included in the initial cost of the SSIT, which is determined by the average cost of implementing an ES for a firm in this industry with a given level of revenues (Mabert et al., 2000; Mabert et al., 2001; Meta Group, 1999). On-going ES expenditures are determined by the students during simulation. In the Investment Decisions component, the ES provides real-time, standardized information that is integrated across the whole firm.
Table 1. Impacts of Enterprise System on a Wireless Telecommunications Firm
The benefits associated with an ES listed in Table 1 were derived primarily from the experience of the expert panel. While specific results for a telecommunications firm are difficult to find in the literature, there are some studies from other industries that corroborate the benefits in Table 1. Robinson (2005) provides anecdotal support from several, mostly manufacturing case studies (see also  ). Hitt et al. (2002) contains results from an extensive study that empirically establishes that firms that adopt ERP will show greater performance. From a survey of business executives, Gefen and Ragowski (2005) show that it is easier for experts to perceive the business value of ERP at a specific IT module level than at an overall enterprise level. This supports our work because our expert panel was only required to assess the benefits at the module-specific level. Overall enterprise value emerges from the simulation exercise.
Perhaps the most relevant work to our current study is found in McAfee (2002), which contains a detailed longitudinal analysis of the benefits derived from an ERP implementation at a manufacturing firm. While the specific metrics used by McAfee do not directly correspond to our metrics, we can make comparisons based on broader categorizations. Specifically, all the percentage improvement metrics in the last column of Table 1 can be categorized as quality and reliability of service improvements (training effectiveness, network quality, average life, and right information) or labor efficiencies (administrative costs, skill level requirements, technician overhead). McAfee (2002) has two metrics that correspond to quality and reliability of service: daily fraction of orders shipped late and daily standard deviation of lead time of orders shipped. Across these two metrics, he reports improvements of anywhere from 34-89 percent which compares to our improvements in the range of 14-100 percent. McAfee (2002) also reports sustained head count reductions of 15-53 percent post-ERP which also compares to our 14-50 percent range of improvements.
The output generated by the simulator with the ES takes advantage of the fact that an ES provides a more integrated environment for data collection and information generation (Davenport, 2000). Therefore, wherever possible, the ES version of the simulator provides more specific and relevant information than the simulator without ES. Additionally, there is no information lag or delay in the simulator with the ES. Table 2 provides a comparison of the type of information generated by each simulator.
Table 2. Output Information Provided in the Simulators of the Firm with and without an ES
To increase confidence that the simulator was working correctly, we ran it through a full suite of widely accepted validation tests (Sterman, 2000).
Critics of simulator-based theoretical and experimental research suggest that the results (whether "good" or "bad") are pre-determined by the equations in the simulator — a criticism that applies to all models, simulated or not. Thus, researchers have the burden of validating and verifying their models. To address this criticism, many researchers opt to use well-known simulators, often accompanying a textbook and used in the classroom, to show that they entered no specific biases (Reagan-Cirincione et al., 1991; Segev, 1987). There is reason to suspect that their results are still a function of the particular formulation of the simulator. With this consideration, our research makes explicit the simulator to be used, along with the underlying assumptions (see Appendix B for a complete listing of the simulator equations). Additionally, we provide a brief overview (see Table 3) of the full suite of well-documented and accepted methods we used to validate and verify the simulator (see Ritchie-Dunham, 2002 for details).
The vast simulation literature provides guidance on what to validate and how, defining validation as confirmation that a computer simulator, within the domain it applies, satisfies the range of accuracy for the application for which it is intended (Forrester, 1961; Sargent, 1999). Validation involves three key components of a simulation: the simulator concept, the simulator operations, and the simulator data (Sargent, 1999).
Simulator concept validity focuses on the reasonableness of the theory and assumptions underlying the simulator. We applied standard concept validation tests (see Table 3) from the system dynamics literature (Sterman, 2000). The simulator passed all the tests, including boundary adequacy, structure assessment, dimensional consistency, parameter assessment, extreme conditions, integration error, behavior reproduction, behavior anomaly, and family membership.
Table 3.System Dynamics Simulator Validity Tests
Simulator operational validity centers on the accuracy of the simulator's behavior, for its intended purpose. We ran the simulator multiple times with random policies, measuring the range of possible responses and outcomes. The expected performance of random allocation should be neutral or negative, since the common belief is that it takes thought to be successful (i.e., make a profit). The results from applying these random policy runs to the simulator seemed reasonable to the expert panel, providing a range of operational believability in the simulator.
Simulator data validity concentrates on the use of adequate and correct data. We obtained data for the initial values of the stocks, parameters affecting the policy within the flow, initial values of the converters, and time series with knowledge from the expert panel and data from two firms in the industry.
This section illustrates how we used the simulator pedagogically in a master's-level course. We conducted the simulation exercise as a two-session module in a master's level supply chain management class in the Red McCombs School of Business at the University of Texas at Austin. The class had 39 second-year master's students, mostly MBA, with the remainder from a joint business/engineering program. Each session lasted 75 minutes. In the first session, the students used the simulator to make strategic supply chain resource allocation decisions for the simulated wireless telecommunications firm every six months over a seven-year period. The second session consisted of a discussion of the results from the simulation exercise and how these results compared to actual company experiences and academic research. While the two-session structure for this exercise may have its downsides (the sessions were actually on different days), it does have the advantage of providing time for the instructor to analyze and organize the simulation results for presentation and discussion in the second session.
As stated in the Introduction, the educational goal of the simulation exercise is to have the students experience the value added of making strategic decisions in a firm's service supply chain with and without an enterprise system. In a pilot test of the simulator, we found students tended to go straight to the simulator without reading the case and instructions. To encourage the students to read the case and instructions and develop their thoughts first, we had the students start the experience, as we describe in Table 4, by opening a web page, where we: (1) used forms to capture demographic data, (2) presented the case study materials seen in Appendix C, and (3) gave the instructions about how to open and run the simulator seen in Appendix D. Then the students clicked a link that opened the simulator that had been installed on their computer. These materials were also available to the students for reference within the simulator. All materials provided to the students are provided in the Appendices, including links to the simulation model, in Appendix E. To promote both focused effort and exploration, the students were asked: (1) to do their best to maximize the value they created for the simulated company, and (2) to explore different strategic hypotheses about how to achieve the best supply-chain-wide consequences. The students were instructed to ask at any time if they had any questions.
At the beginning of this exercise, all students received the same information. Specifically, they were given details about the case and simulation procedures. No prior preparation on ES was given. Students worked individually. The students started by entering demographic information about themselves, for example, work experience. Several screens in the simulator walked them through the case description and exercise instructions (see Appendices C and D). Each student was randomly assigned to either the with ES or without ES simulation scenarios. After this, they ran the simulation, in which they were provided the information shown in Table 2 and allowed to allocate financial resources to: (1) ordering base stations, (2) hiring human resources, (3) firing human resources, (4) training human resources, and (5) acquiring service supporting information technology. The debriefing was conducted in a subsequent session. A flow of the entire exercise is given in Table 4. Table 4 also includes the time it takes per section of the exercise and the materials required. These times were validated during the exercise by recording the actual time taken by each participant in each section.
Table 4. A Flow of the Simulation Exercise
We tracked the economic value created by each student in the simulator using the widely used measure Market Value Added (MVA) (Ehrbar, 1998). Although each scenario ran for seven simulated years, we used the MVA at the end of the fifth year as the performance measure in order to avoid end game effects.
The debriefing session opened with a general discussion about the students' experience with the simulation exercise. From their comments and questions, we found that the students were most interested in: (1) mapping the ES impacts on the business flow of materials and information; (2) understanding the influence of these ES impacts on organizational performance; (3) understanding how ES influence the quality of strategic decisions; (4) seeing the whole firm and its supply chain; (5) focusing on a service application; and (6) the learning approach (i.e., using a computer simulation exercise). The issues surfaced in the discussion we led in the debriefing session and from comments on written evaluations.
Next, we focused on the student's experience of the simulator exercise. What did they notice? What seemed to influence their performance? What did they want to achieve and how were they helped or hampered by the information they were provided?
We have noticed common responses to these questions. The students typically notice that: (1) it is harder to create value for the firm than they thought after reading the case study; (2) delays in investing in infrastructure slow down market growth; and (3) they have to invest in infrastructure, human resources, and training to keep customers satisfied, so they can maintain market share with the rising market demand. They tend to notice that their performance is most influenced by: (1) when they invest in infrastructure; (2) the information they have about what is going on; and (3) how much they sustained their investment in growth of infrastructure and human resources. When discussing what they wanted to achieve, two camps emerge, focusing on either: (1) rapid growth of economic value added, at all costs; or (2) sustainable growth in financials and customer satisfaction. They typically comment that having what appeared to be real time data helped see the consequences of their actions more clearly, while seemingly delayed information hampered seeing these consequences. Also, having integrated information helped them see the linkages across the supply chain, versus seeing only what was happening in each function.
We then present the statistical results to compare their experiences of having or not having an ES. Here we focus on comparing the experiences and results the students had in the two groups ("with ES" and "without ES"). Results are presented in the form of simple descriptive statistics and as a regression analysis. The regression analysis included fifth year MVA as the response, the simulator scenario as a predictor, and a number of other factors controlling for different types of work experience amongst the participants (e.g. years of managerial work experience, and years of experience in the telecommunications industry). We explore whether the differences in results between the groups make sense and why. Note, while it may seem desirable to have each student do both scenarios (i.e., with and without an ES), we have found that it is not an absolute necessity when this type of post game comparison of the results is used because patterns are quite likely to emerge from the aggregated data in the statistical analysis. A single scenario per student is not only more economical in terms of time, it also avoids additional complications such as the need to correct for a learning effect.
The analysis and comparison of the results segues nicely into a conversation about the impacts of an ES on a firm, discussing the impacts of ES on business flows of materials and information, organizational performance, strategic decision making quality, and seeing the whole firm and its supply chain. Here we introduce a discussion on studies found in the academic literature and the popular press. This includes a discussion of the mixed results and possible reasons for these mixed results. This discussion includes negative ES impacts on business performance, such as the loss of flexibility, loss of competitive advantage, training difficulties, cost overruns, and other implementation difficulties (Griffith, Zammuto, and Aiman-Smith, 1999; Hong and Kim, 2002; Songini, 2004). Additionally, we discuss the results emerging in carefully designed ES-impact studies (Hitt et al., 2002; McAfee, 2002), that find positive business performance from better systems integration, standardization of data and processes, end user visibility across the business enterprise, and improved decision support functionality. Such a discussion provides the opportunity to talk about how hard it is for researchers to assess the impact of ES on business performance from empirical data sets relative to a carefully designed simulation experiment. We also take the opportunity to discuss complexities not captured by the simulation exercise such as the learning curve effect documented by McAfee (2002).
We have presented a simulation exercise to help students assess the impact of a successfully implemented ES on the service supply chain of a telecommunications firm. We have found that the simulation experience and the subsequent debriefing provide a valuable learning experience for both students and instructors. Our overall experience with such exercises is that the students find them to be of great educational value (Anderson and Morrice, 2000; Bates, 2002).
The over-arching principles that guided the development of this simulation exercise were instructional brevity and ease of use because we only have a limited amount of time to cover such topics in our supply chain and information management courses. From discussions with other instructors, executive education consultants, and textbook authors, we recognized that this was not unique to our program. The exercise requires a standard PC with Microsoft Windows, an executable version of the simulator and an isee Player from ISeeSystems to run iThink simulation models (see Appendix E).
As we mentioned, the simulation exercise can be completed in three to six hours of class time. We presented a three-hour example of a simulation exercise that works well in our supply chain management course. Many other configurations are possible. For example, with more class time, each student could participate in multiple simulation scenarios possibly providing an even richer educational experience. We have done such things with other simulation exercises. However, when students perform more than one simulation, care must be taken to account for learning effects (and the corresponding statistical biases) if the simulation results are to be used in a subsequent debriefing session (Anderson and Morrice, 2000).
Although the statistical results discussed in Section 5.2 are rich in educational value and provide a basis for excellent discussion, much more could be done with the data collected from the simulation exercise. The simulator tracks the performance of each student over time. Analysis of this data provides the opportunity to discuss dynamical decision making behavior and how this impacts business performance. There is a strong and growing literature in this area due to the interest in supply chain dynamics (Sterman, 2000). Based on our experience, we do need to add one word of caution at this point. Including a statistical analysis of the simulation dynamics requires more time, care, and statistical sophistication on the part of the instructor and the students (it generally requires some knowledge of time series analysis). Nevertheless, this analysis can be valuable because it often helps to explain the patterns observed in the more "static" statistical analysis mentioned in Section 5.2.
We thank the AE and three reviewers for feedback that significantly increased the teaching value of this paper. Our research was supported in part by a grant from the SAP America University Alliance program, a fellowship from the University of Texas at Austin, software from ISee Systems (previously High Performance Systems), and funding from the Institute for Strategic Clarity.
Akkermans, H.A., P. Bogerd, E. Yucesan, and L.N.v. Wassenhove (2003), "The Impact of ERP on Supply Chain Management: Exploratory Findings from a European Delphi Study," European Journal of Operational Research, Vol. 146, pp. 284-301.
Andersen. (1999), Enterprise Business Solutions: The Andersen Consulting Survey of Chief Executive Officers: Andersen Consulting Institute for Strategic Change.
Anderson, E.G., and D.J. Morrice (2000), "A Simulation Game for Service-Oriented Supply Chain Management," Production and Operations Management, Vol. 9, pp. 40-55.
Bates, J. (2002), "Littlefield Technologies: I Love This Game!," Texas Business Weekly, , (last accessed on November 6, 2002).
Davenport, T.H. (2000), Mission Critical: Realizing the Promise of Enterprise Systems, Harvard Business School Press, Cambridge, MA.
Ehrbar, A. (1998), EVA: The Real Key to Creating Wealth, John Wiley & Sons, Inc., New York.
Ettlie, J.E. (2000) "Technology and Weak Appropriation Conditions: The Adoption of Resource Planning Systems," Paper presented at the Institute for Operations Research and Management Science, Salt Lake City, Utah.
Forrester, J.W. (1961), Industrial Dynamics, Productivity Press, Cambridge MA.
Gattiker, T., and D. Goodhue (2000) "Understanding the Plant Level Costs and Benefits of ERP: Will the Ugly Duckling Always Turn into a Swan?," Proceedings of the Thirty-Third Annual Hawaii International Conference on System Sciences.
Gefen, D., and A. Ragowsky (2005), "A Multi-Level Approach to Measuring the Benefits of an ERP System in Manufacturing Firms," Information Systems Management, Vol. 22, No. 1, pp. 18-25.
Griffith, T.L., R.F. Zammuto, and L. Aiman-Smith (1999), "Why New Technologies Fail," Industrial Management, No. May-June, pp. 29-34.
Hitt, L.M., D.J. Wu, and X. Zhou (2002), "Investment in Enterprise Resource Planning: Business Impact and Productivity Measures," Journal of Management Information Systems, Vol. 19, No. 1, pp. 71-98.
Hong, K.-K., and Y.-G. Kim (2002), "The Critical Success Factors for ERP Implementation: An Organizational Fit Perspective," Information & Management, Vol. 40, pp. 25-40.
ISeeSystems (2001) Ithink (Version 7.0.1), Hanover, NH.
Mabert, V.A., A. Soni, and M.A. Venkataramanan (2000), "Enterprise Resource Planning Survey of US Manufacturing Firms," Production and Inventory Management Journal, Vol. 41, No. 2, pp. 52-58.
Mabert, V.A., A. Soni, and M.A. Venkataramanan (2001), "Enterprise Resource Planning: Measuring Value," Production and Inventory Management Journal, Vol. 42, No. 3/4, pp. 46-51.
Markus, M.L., and C. Tanis. (2000) "The Enterprise Systems Experience -- from Adoption to Success," in Framing the Domains of IT Management: Projecting the Future ... Through the Past, ed. R.W. Zmud, Pinnaflex Educational Resources, Inc., Cincinnati, OH.
McAfee, A. (2000), "The Impact of Enterprise Information Systems on Operational Effectiveness: An Empirical Investigation," Ph.D. Thesis, Harvard Business School, Boston, MA.
McAfee, A. (2002), "The Impact of Enterprise Technology Adoption on Operational Performance: An Empirical Investigation," Production and Operations Management, Vol. 11, No. 1, pp. 33-53.
Meta Group (1999, October), "ERP Implementation Study Reveals Costs, Benefits," APICS -- The Performance Advantage.
Morecroft, J. (1983), "Concepts, Theory, and Techniques: A Systems Perspective on Materials Requirement Planning," Decision Sciences, Vol. 14, No. 1, pp. 1-18.
O'Leary, D. (2000), Enterprise Resource Planning Systems: Systems Life Cycle, Electronic Commerce, and Risk, Cambridge University Press, New York.
Reagan-Cirincione, P., S. Schuman, G.P. Richardson, and S.A. Dorf (1991), "Decision Modeling: Tools for Strategic Thinking," Interfaces, Vol. 21, No. 6, pp. 52-65.
Ritchie-Dunham, J.L. (2002), "Balanced Scorecards, Mental Models, and Organizational Performance," Ph.D. Thesis, Department of Management Science and Information Systems, University of Texas at Austin, Austin, Texas.
Ritchie-Dunham, J.L., D.J. Morrice, E.G. Anderson, and J. Scott. (2000) "A Strategic Supply Chain Simulation Model." in Proceedings of the 2000 Winter Simulation Conference, eds. J.A. Joines, R.R. Barton, P.A. Fishwick & K. Kang, Institute of Electrical and Electronics Engineers, Piscataway, New Jersey.
Robinson, P. (2005), Business Excellence: The Integrated Solution to Planning and Control, BookSurge Publishing, North Charleston, SC.
Sargent, R.G. (1999) "Validation and Verification of Simulation Models," in Proceedings of the 1999 Winter Simulation Conference, eds. P.A. Farrington, H.B. Nembhard, D.T. Sturrock & G.W. Evans, Institute of Electrical and Electronics Engineers, Phoenix, Arizona.
Segev, E. (1987), "Strategy, Strategy-Making, and Performance in a Business Game," Strategic Management Journal, Vol. 8, pp. 565-577.
Senge, P.M. (1990), The Fifth Discipline, Doubleday Currency, New York.
Sharda, R., S. Barr, H. McDonnell, and C. James (1988), "Decision Support System Effectiveness: A Review and an Empirical Test," Management Science, Vol. 34, No. 2, pp. 139-159.
Songini, M.L. (2004, August 23),"Ford Abandons Oracle Procurement System: Switches Back to Mainframe Apps.," Computerworld, 7.
Sterman, J.D. (2000), Business Dynamics: Systems Thinking and Modeling for a Complex World, Irwin McGraw-Hill, Boston.
|
Volume 7, Number 3, May 2007
