Chapter 12 Selecting the Best Alternative Design Strategy



Chapter 11

Selecting the Best Alternative Design Strategy

Chapter Overview

The primary purposes of Chapter 11 are to show students that all of the software associated with a systems development project is not developed in-house; emphasize that analysts should consider several design strategies before choosing one to pursue for further development in design; and return to and update the Baseline Project Plan first developed during project initiation and planning (Chapter 6). This chapter’s secondary purpose is to emphasize that the consideration of a packaged software solution should be done after the analysis efforts are complete, not as a substitute for analysis.

The authors discuss the various sources of software and the reasoning that should be followed when choosing among the many options available to an analysis team for developing design strategies. The point is that the make versus buy decision is not a choice of one or the other, but is in reality a spectrum of choices ranging from make at one end and buy at the other. Just as important, more choices these days are made toward the buy end of the scale. The authors include a discussion of outsourcing, an option for systems development and management that may not occur to many students in their first systems development course.

When discussing design strategies (which the authors define as “a high-level statement about the approach to developing an information system, which includes statements on the system’s functionality, hardware and system software platform, and method for acquisition”), the authors revisit Hoosier Burger’s inventory control system. The authors show how, given its system objectives and constraints, Hoosier Burger develops a series of design strategies for its new system. The authors present a weighted multi-criteria decision-making matrix as one method for comparing design strategies side-by-side and selecting one alternative for further development. Instructors may want to demonstrate design strategies that are not presented in this textbook.

In the later part of the chapter, the authors revisit and update the Baseline Project Plan from Chapter 6. Updating an entire Baseline Project Plan (BPP) is beyond the scope of this chapter, so instead the authors list the sections of the BPP that are updated and show how more accurate information is used to update both the cost benefit analysis and the project schedule. The BPP should be updated at this point, since the end of the analysis phase typically is a project milestone, often with review by those that must approve the continuation of project funding.

Instructional Objectives

Specific student learning objectives are included at the beginning of the chapter. From an instructor’s point of view, the objectives of this chapter are to:

1. Show students that in-house developed systems are not the only source for software.

2. Teach students how to assemble the various pieces of an alternative design strategy.

3. Demonstrate how to generate three alternative design strategies for an information system, focusing on what should and should not be part of the strategy at this stage and why.

4. Show students one method for choosing among design strategy alternatives.

5. Discuss alternative generation and design strategy selection for Internet-based electronic commerce applications.

Classroom Ideas

1. Use Table 11-2 and an updated version of the same information from the most recent Software Magazine survey to begin a discussion of the many, varied sources of software in the marketplace.

2. Use Table 11-5 to summarize the alternative sources for software and how to choose among them for specific software needs. This table can serve as the basis for a discussion of the make versus buy decision and can be expanded to include the “not invented here” syndrome.

3. Emphasize the concept of a design strategy to your students. A design strategy is a relatively difficult concept for many students to grasp. Their first inclination is to think of a strategy as simply a set of system functional requirements, and that different strategies are many, moderate, and few (basic) functions. Certainly, this is one component of a design strategy, but point out that operational platform and acquisition source are also choices that must be made to set the direction for design efforts.

4. Have students research the proper format and contents for Requests for Proposal and have them create and/or present RFPs for a specific design strategy (see Problem and Exercise 2). RFP preparation should include discussion of the hardware, software, and organizational issues presented in this chapter.

5. Work through the Hoosier Burger example in class, including creating the multi-criteria decision matrix. Expand student understanding of this comparison technique by introducing changes in Hoosier Burger’s objectives and constraints and showing how these changes affect the matrix and maybe the final choice. Alternatively, give students another example, with objectives, constraints, and some design strategy options. Ask your students to generate alternative strategies and the corresponding matrix. Problems and Exercises 9 and 10 are good problems for students to work in class.

6. Using the Hoosier Burger multi-criteria decision-making matrix (Figure 11-5), show how changes in existing weights and ratings affect the alternative strategy scores and the final choice. Lead a class discussion on how these numbers are generated and what these numbers really mean. Ask students how they think management will actually use the numbers provided in the matrix. For example, what happens if management really supports one of the design strategies that was not chosen as optimal?

7. In class, compare and contrast Hoosier Burger’s multi-criteria decision matrix with other decision-making techniques (such as elimination by aspects, allocating points out of some total, or ranking). Have students judge when each technique is preferable to others and why. Discuss the concept of mandatory, essential, and desirable features or criteria. Discuss how to apply these three kinds of features/criteria under different multi-criteria decision-making techniques.

8. In class, work through the Baseline Project Plan and the changes that were discussed in the chapter, using Figures 11-6 through 11-10 and Tables 11-6 and 11-7. Have students make some of the specific changes that are discussed more abstractly in the text. Discuss with students the reasoning behind a Baseline Project Plan and the purpose it serves. (At this point, you may want to assign Problem and Exercise 7.)

9. Have students identify an Internet-based electronic commerce application, perhaps one that you have discussed in previous class periods. For this application, have students investigate different design strategies.

Answers to Key Terms

Suggested answers are provided below. These answers are presented top-down, left to right.

| 8. Web server | | 5. Outsourcing |

| 3. Design strategy | | 6. Request for proposal |

| 7. Scalable | | 1. Application server |

| 4. Enterprise resource planning | | 2. Application service provider |

Answers to Review Questions

1. The deliverables from generating alternatives and selecting the best one include: (1) identifying at least three substantively different system design strategies for building the replacement information system; (2) choosing the design strategy judged most likely to lead to the most desirable information system; and (3) preparing (updating) a Baseline Project Plan for turning the most likely design strategy into a working information system.

2. Analysts generate three alternatives because three alternatives can neatly represent both ends and the middle of a continuum of potential solutions.

3. Six sources of software are identified in the text. These include hardware manufacturers, packaged software producers, custom software producers, enterprise-wide solutions, application service providers, and in-house developers. Hardware manufacturers are among the world’s leading software producers, but the software they make is limited mostly to operating systems and utilities. In contrast, packaged software producers develop all kinds of software for all kinds of markets. Custom software producers, sometimes called consulting firms, develop specific, customized software that matches a client’s particular needs. Enterprise solutions consist of a series of integrated modules; these modules are integrated to focus on business processes rather than on business functional areas. Application service providers are organizations that host and run computer applications for other companies, typically on a per use or license basis. In-house development requires the resources, especially trained staff, to develop software targeted to an organization’s own specific needs.

4. When deciding what off-the-shelf software to buy, you should compare products and vendors. Additional criteria include (among others that are more situation-specific): cost, functionality, vendor support, vendor viability, flexibility, documentation, response time, and ease of installation.

5. When trying to determine if new hardware or system software are necessary, analysts focus on whether a particular design strategy can be run on existing hardware and systems software platforms. Advantages of running the target system on existing platforms include lower costs, familiarity, ease of integration, and few or limited data conversion costs. On the other hand, parts of the target system may require specific platforms, new platforms allow for substantial upgrading, and a new platform may allow the organization to radically change its computing operations. An RFP is a formal document that provides detailed specifications about a target information system and asks vendors for information on how they would develop the system. Analysts use RFPs as a way to get vendors to perform the necessary research into specific design strategies and the hardware and systems software vendors believe are necessary for developing the new system.

6. Other than hardware and software, issues analysts must consider include implementation issues (such as changes in work relationships, training, and disruptions in work procedures) and organizational issues (such as cost of implementation, determining what management will support, and whether users will accept the new system).

7. Analysts consider many issues in developing alternative solutions to information system problems. Of particular interest are the system owner’s and users’ prioritized system objectives and system (and development) constraints. Analysts consider which design strategies would minimally satisfy objectives and not violate constraints, on the one hand, as well as which design strategies would meet or exceed objectives with minimal violation of constraints. Many possible design strategies exist between these two extreme positions.

8. While alternative design strategies are compared in many different reasonably objective ways, the chosen design strategy depends on what management’s true objectives are for a particular development project. Management may ignore constraints, or alternatively, choose the least expensive system to develop, regardless of which design strategy appeared to be the best in the objective comparison.

9. Every section of the Baseline Project Plan is updated during the alternative generation and selection step of analysis.

10. To verify vendor claims about a software package, an analyst can ask for a software demonstration, use the software (and its documentation and training materials) personally, talk with other users of the software, and consult independent software testing and abstracting services (surveys available for a fee).

11. Enterprise resource planning systems consist of a series of integrated modules; these modules pertain to specific, traditional business functions. However, these modules are integrated to focus on business processes rather than on business functional areas. Enterprise resource planning systems advantages include a single repository of data for all aspects of a business process, flexible modules, less maintenance, more consistent and accurate data, and ease of adding and integrating new modules into the existing system. Possible disadvantages of this approach include complexity, lengthy implementation, lack of in-house expertise, expense, and changing how the organization conducts its business.

Answers to Problems and Exercises

1. In addition to visiting Datamation’s Web site, encourage your students to also visit Software Magazine’s Web site. Tables 11-2, 11-3, and 11-4 summarize information from Software Magazine. The most recent Software 500 list is located at .

The software industry continues to be a high-growth, dynamic sector of the economy. Thus, students are likely to find changes in the more recent rankings. Microsoft continues to grow and increase its dominance. There have also been some recent mergers, which will affect the rankings. We see changes in the rankings as companies sell more or less software. Some software vendors have done a good job of forecasting market needs and then producing and selling good products to meet those needs. Two examples of this are Microsoft, with its Windows-based products, and Lotus, with Notes and its bevy of other applications.

2. An organization uses the Request for Proposal (RFP) to solicit proposals from several competing vendors. Usually, RFPs first provide background information on the company and the business units involved in the request, an explanation of the information systems needs, a description of what is wanted from the vendors (i.e., what information they must provide or other actions they must take), and an explanation of any rules or procedures for the RFP and system development process. The bulk of the document then describes the mandatory, essential, and desirable requirements in the areas of need (e.g., functionality, hardware, software, and service). Students’ RFP outlines should include these key features.

3. In the first accompanying spreadsheet, the weights have been changed to reflect a weighting that represents systems development projects in governmental agencies. In these settings, a great deal of weight is given to the costs and time frame of the system being proposed. As often happens in governmental agencies, by weighting costs and timing more heavily in this example the outcome has changed. Alternative B and Alternative C have traded places; Alternative B now has the highest score.

The second accompanying spreadsheet reflects changes in the ratings for each alternative. Alternative B now has a much higher score. This shows how important ratings are and how subtle changes and/or biases in ratings can significantly impact the final outcome.

The third spreadsheet has additional criteria of “references” and “in use.” “References” are written references from other organizations currently using the vendor’s systems. “In use” means that the technology proposed has been in use by a paying customer for 6 months. These are very common criteria for organizations that value stability of the vendor and technology over performance and state-of-the-art technology. These additional criteria do not change the outcome; Alternative B still has the highest score.

4. The list for evaluating computer hardware and system software is similar to the list for selecting off-the-shelf application software. In addition to cost, functionality, vendor support, vendor viability, flexibility, documentation, response time (or other performance criteria, such as number of transactions per unit of time, other benchmark tests, or raw processor speed), and ease of installation, we might include the current staff’s familiarity with the new system, need for retraining, compatibility and connectivity with current systems, cost of converting old systems to the new platform, ease of future upgrades, and the track record of the vendor in successfully implementing this new system in other organizations.

5. For this method of evaluation to be valid, one assumes that all relevant criteria are known and included and that the weights and ratings are accurate. More important, this method assumes that the alternatives and criteria lend themselves to a quantitative analysis. Some people argue that some parts of the analysis inherently cannot be quantified. For example, it is difficult to truly explicitly rate and weight the amount of trust that you can place in a vendor and your belief that he will indeed follow through on his claims. Further, this method assumes that the criteria are independent of each other, and thus scores are additive.

6. Another quantitative method for choosing between alternative systems is to choose almost exclusively based on cost. In most state and federal agencies, information systems are chosen this way. People in these agencies are often forced by law to choose the least expensive system that meets some minimum level of adequacy for the relevant information system needs. Indeed, many states and parts of the federal government are attempting to change this method of awarding governmental contracts. A second, more qualitative method chooses the vendor that supplies the necessary system and also provides the best value-added services and components. For example, many companies will choose a vendor that, in addition to the basic system, is willing to supply additional training, free components, less expensive maintenance, or has offered to enter into a strategic partnership that is of mutual benefit to both companies.

When multiple decision makers collaborate and make the decision together, several methods for incorporating their interactive input exist. For example, for a quantitative analysis, such as that presented in Figure 11-5, each decision maker enters his ratings for each alternative across each criteria. This data are then summarized using a spreadsheet, and the average and/or summary rating across all decision makers is used to choose a vendor. In addition, group support systems can tally the individual, anonymous ideas, comments, and/or votes of multiple, collaborating decision makers. Since many multi-criteria decision-making methods exist (e.g., another popular technique uses pairwise comparisons), students will identify several alternatives for this question.

7. Basically, your students should include the deliverables from the three major sub-phases of analysis: requirements determination, requirements structuring, and alternative generation and selection. Key issues include meeting participants, the level of detail to present to Bob Mellankamp, and, subsequently, how long the meeting should last. The meeting should not only summarize the findings from analysis but also validate these findings by outlining the process followed. An updated BPP should be covered. Without more information, it may be difficult for the students to decide on other key issues. Have your students identify the information that is needed to decide on these issues. For example, they may want to know more about Mellenkamp’s personality and preferences before making a decision.

8. In addition to cost, functionality, vendor support, vendor viability, flexibility, documentation, response time, and ease of installation, a number of other “real world” criteria might be included. People often choose application packages, such as word processors and spreadsheets, based solely on their familiarity with the packages and/or their bias toward one hardware platform or operating system over another. To a certain extent this is functional. On the other hand, this can be dysfunctional. For example, it is useful to consider the current staff’s familiarity with the new application software and the resulting need for retraining. However, it is dysfunctional if a company does not choose new software because of the employees’ lack of familiarity with the software. Eventually, software will evolve and the market will change, and the employees (and the company) will be left behind using antiquated technology. Additional criteria include compatibility with currently used application software (so, for example, data can be shared), compatibility with existing hardware and system software, ability to support a range from novice to experienced (or power) users, and appeal of the user interface (ease of use).

9. Some basic system alternatives include: (1) build the system in-house using a programming language such as C or Microsoft Visual Basic; (2) purchase off-the-shelf packages, such as Microsoft Excel and Access, along with the necessary computer and telecommunications equipment, and use these to build the necessary systems in-house; (3) purchase custom software and a turn-key system from either a generic information systems consulting firm such as Andersen Consulting or EDS or from a specialized systems provider for the food retail or pizza retail industry; or (4) outsource the systems to an outside firm.

Each of these alternatives has unique advantages and disadvantages. Given that the system needs to be stable, easy to use, and built relatively quickly, and given that this organization probably has little or no in-house systems personnel, the third or fourth alternatives are most realistic.

10. It is useful for students to create a spreadsheet like the one presented in Figure 11-5. Your students should list in the criteria category each of the requirements and constraints described in the problem, plus any other criteria that they believe are relevant. As best they can, have your students weight each criterion and use this information to rank each of the four alternatives presented in the previous answer. Your students will need to make some assumptions in order to complete each of the ratings. This method of analysis is useful even for this relatively small system. This method forces the decision maker to flesh out relevant criteria and weighting and be as objective as possible in rating each alternative on each criterion.

11. Having only one design strategy is problematic in several ways. First, there is no guarantee that the best, correct, or even an adequate system for the situation is being developed or purchased. This is not obvious because it is unclear if other alternatives were considered, and if they were, those present cannot see why the one choice won out. Second, if the one strategy is chosen because only one vendor is used, there are no benefits from having multiple vendors compete for an RFP. For example, the vendor has no incentive to keep his price as low as possible. Third, without the detailed, public systems specifications that are part of a competitive bid process, there is not likely to be much in the way of written documentation to refer back to if the vendor does not fulfill his promises. If the analysts present only one design strategy during the oral presentation to the project steering committee or client, the recommendations are likely to be (at worst) rejected, or (at best) accepted with great skepticism. It is also possible that those present at the meeting will start to generate alternatives, each representing that person’s position. The meeting will quickly deteriorate since a fair assessment of ad hoc alternatives cannot be done within the limits of a meeting. In any event, this is not a good way to begin the development of an information system (or build a career).

12. The list for evaluating alternative custom software developers is similar to that for selecting off-the-shelf application software or for computer hardware and system software. In addition to cost, functionality, vendor support, vendor viability, flexibility, documentation, response time, and ease of installation, you might include the current staff’s familiarity with the software, need for retraining, compatibility and connectivity with current systems, and the track record of the vendor in successfully implementing similar software in other organizations. Such vendors should have an established track record for developing similar software in other organizations. Their references should be checked thoroughly, including visits to other sites. If the developer’s role ends after the application is accepted, then the vendor’s reputation for handling this transition from external development to internal maintenance is important. From a legal point of view, you may want to select a custom developer based on the willingness to sign a non-disclosure agreement, so that he is not allowed to develop a similar system for one of your competitors, at least for a certain amount of time.

13. The project team can use the advantages of the enterprise resource planning design as part of its strategy for selling this system. The team can stress that this solution consists of a series of integrated modules; these modules are integrated to focus on business processes, and the firm can integrate all parts of a business process. This approach includes a single repository of data, thus providing more consistent and accurate data and less maintenance. These modules are flexible and new modules are easily integrated into an existing system.

Also, an enterprise resource planning (ERP) system might be justified on the following grounds: (1) it is a complete enterprise-wide solution that models all aspects of each transaction, supposedly seamlessly and within a single system; (2) an ERP system is based on a single repository of all corporate data, which implies consistency, accuracy and flexibility of the data; and (3) adding new modules should be relatively painless as all modules are specifically designed to work together. On the other hand, some might counter that ERP systems are complex and require expensive outside expertise for implementation, and that the organization itself has to adjust to fit the software's model of how organizations should operate rather than the reverse.

Guidelines for Using the Field Exercises

1. Since pricing and capabilities change rapidly, ask your students to visit the Web sites of several vendors, including Compaq, IBM, Gateway, and Dell. For discussion purposes, three alternatives are presented below; however, please note that the following information should be updated each semester.

Low-end alternative: 866MHz Pentium III, 128MB of RAM, 10GB hard drive, 48X CD-ROM, 15” color monitor, running Microsoft Windows 98, and equipped with Microsoft Office 2000. Cost is about $1,100.

Mid-range alternative: 1.4GHz Pentium IV, 128MB of RAM, 20GB hard drive, 17” color monitor, a 48X CD-ROM, running Microsoft Windows ME, and equipped with Microsoft Office 2000. Cost is about $1,700.

High-end alternative: 1.7GHz Pentium IV, 256MB of RAM, 60GB hard drive, 21” monitor, DVD-ROM drive, running Microsoft Windows 2000, and equipped with Microsoft Office XP Professional. Cost is about $2,400.

2. The spreadsheet below presents the quantitative analysis for the three alternatives described in the previous answer. The criteria are speed, storage, ease of use, reliability, costs, and time to operation. The mid-range alternative has the highest score because it has acceptable rankings for the performance-oriented criteria and scores well on the highly weighted criteria of costs. As is often the case, people often buy more technology than they really need. The counter argument is that, for PCs, one should buy as much power as they can reasonably afford so the technological life of their equipment is longer, and they will be better able to take advantage of new software as it becomes available.

3. Students should easily locate three different spreadsheet products and readily obtain related information about each product. Your students might also visit a computer retailer, particularly a large one such as CompUSA or Egghead Software. At such stores the students can test a variety of software packages. Some of the criteria the students are likely to use include speed, ease of use, compatibility with other packages, price, and features. It is useful for students to compare their analyses. This comparison helps them understand how difficult it is for software vendors to be all things to all customers.

4. Business people are likely to use this list of criteria in some formal or informal way. Have your students present their findings to the class so that they can learn about a variety of companies. It is useful for students to see how companies actually use these criteria and methods in purchasing off-the-shelf software.

5. Obtaining RFPs may be difficult for your students. It may be necessary for you to obtain copies of RFPs either from business contacts or from the university. It is very useful for students to see real RFPs. Students are amazed at how lengthy and detailed these are for larger, more complex systems. Your students are likely to be amazed at how complicated RFPs can become for governmental agencies.

6. Chances are any organization that a student contacts about its ERP implementation will have a lot to say about it, provided the IT staff is willing to share the information. Typically, ERP implementations take several years and cost quite a bit in terms of consultant fees. There are many reasons to move to an ERP system, just as there are many reasons not to move to an ERP system. The organization was probably attracted to the promise of uniformity and consistency made by ERP vendors, although the exact reasons differ from firm to firm. Chances are good that the organization has made some internal changes, such as realigning departments internally to take advantage of the opportunities ERP systems offer, as well as to meet the demands ERP systems make in order to operate effectively. The chances are also good that most of the implementation work has been done by outside consultants, so for firms not used to managing large numbers of contractors, an ERP implementation is a new and different experience. The implementation is likely still going on at whatever firm a student happens to talk with, but it has probably been going on for many months or years, as each ERP implementation is a learning experience for the consultants and the adopting firm.

Guidelines for Using the Broadway Entertainment Company Cases

Guidelines and answers for using the Broadway Entertainment Company cases are available on this textbook’s companion Web site. Please visit hoffer to access this information.

-----------------------

Error! Not a valid link.

Error! Not a valid link.

Error! Not a valid link.

Error! Not a valid link.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download