 Implementation Guide for Electronic Commerce  Table of Contents
Introduction Overview of Guide This section of the MIRL EC site is an operational guide to implementing electronic commerce. In it we provide instructions about what to do, and when. This guide borrows heavily from the primer and common questions, but adds both a great deal of information and a strong emphasis on operations. Sections are organized in terms of a five-stage EC technology implementation life cycle. 1. Awareness Awareness provides the beginning of a consensus for change, an understanding of possibilities, and a common framework for discussion and planning EC. 2. Business analysis The only reason to implement EC is to improve one’s business. Thus it is critical to understand the relationship between EC and business needs. Certain understanding of this relationship is impossible because in the real world business conditions and technological possibilities are constantly changing. Further, the implementation of EC may create business possibilities that did not previously exist. Still, business goals can be defined and uncertainties can be reduced. The better the business analysis, the greater the probability of high returns from an investment in EC. 3. Requirements definition "Requirements definition" is identification of the specific EC system that will meet the previously defined business needs. The goal is to get to the point of being able to commit to implementation. To make such a commitment you need to know: 1- what technology and business process you need to meet your business goals, and 2- how much they will cost. 4. Implementation and validation The purpose of the implementation phase is to acquire and implement the system. This is the phase when new technology comes in the door, training is conducted, reporting relationships change, and new EC processes begin to function. As components are installed, validation testing is needed to make sure that the system performs in accordance with its specifications. Does it do what it is supposed to do, and not do what it is not supposed to do? Success at this stage requires the ability to manage both technological and organizational change. While the two are inextricably linked, we separate the them here for purposes of explanation. A common example is used to demonstrate the interactions between them. 5. Maintenance and evolution Once testing provides enough confidence in the system, it is "deployed", i.e. it is used for routine business activities. As daily use proceeds, the workings of the system are continually assessed in terms of system failures, evolving expectations, new business needs, and new technological possibilities. There is a great deal of overlap and interaction among these stages, and particularly between adjacent stages. Another objective of this guide is to show the dynamics of these overlaps, and how they can be managed in the service of powerful EC implementation. Finally, this guide employs a single example (Jackson Windows) for all sections, and builds on that example at each stage. The structure of this Guide can be visualized as in Figure 1. Common Example: Jackson Windows Jackson Window is a medium sized family run company. It has been making wood windows for the residential and commercial construction market for many years. Annual volume is about $20 Million and the company employs about 180-200 people. It does business from a single facility of about 100,000 ft2 . Much of the square footage is being used for inventory storage. Forty years ago, when ‘Dad’ started the business, Jackson produced a few window types in large quantities. Business was good, and during the construction season Jackson couldn’t keep up with the demand for its windows. So the company developed the habit of building up their inventory during the off-season to accommodate demand during the busy season. Over the years, Jackson’s customers began asking for more and more options and features on their windows. Glass offers a good example of these options. Originally, Jackson sold windows with only single panes of glass. As people became more energy conscious, they began to demand thermal pane sealed units. But of course, since these cost more, customers didn’t want to put this type of window in a garage or storage shed, which were unheated anyway. So Jackson had to supply windows with both single panes and sealed units. From a marketing point of view, this is easy. You simply offer customers a choice of glass type. But the way Jackson manufactures windows made that exceedingly difficult. Essentially, Jackson starts with a piece of glass and builds the wooden window around this piece of glass. So the only way Jackson could still ship from stock and respond to the demand for different glass was to stock both type of windows, some with a single pane and some with sealed units. This increased inventory. Subsequently some customers wanted sealed units with Argon gas, which is even more energy efficient. But now Jackson had to stock 3 types of windows. So as Jackson added new features, inventories ballooned. Currently inventories are very, very high and are threatening to increase even further. Furthermore, Jackson never seems to have the right products on hand to satisfy customer demand. Somehow, in spite of a large inventory, what is needed never seems to be in stock. Predicting customer demand is very difficult especially in the residential construction market. As a result, customers are complaining about long lead times and missed delivery dates. Clearly, something needs to be done to remedy this situation. Management decided to make the following changes:
Awareness
Outline
What is EC, and why is it different from other forms of commerce? Electronic commerce is the application of communication and information sharing technologies among trading partners to the pursuit of business objectives. In the real world of business four distinct types of EC mingle and interact: Information access provides search and retrieve capability for public domain and proprietary data archives. The most common example of this type of EC is information services (e.g. Compuserve, AOL, Nexus - Lexus, and ABI Informs) that maintain data bases and charges for access. Another example would be a large manufacturer communicating engineering change notices to suppliers via dial-up access to a centralized database. Interpersonal communication services provide methods for parties with mutual interests to exchange information, "discuss" ideas, and improve their cooperation. Examples include customer and supplier design groups jointly working out product specifications, updated files being sent by a publisher to a printer, and a purchasing agent using electronic mail to negotiate an expediting schedule with a supplier. Shopping services allow people to seek and purchase goods or services though electronic networks. This form of EC for retail sales is what comes most readily to mind when one hears the term "electronic commerce". But this genre of EC can be extended in many other directions. As examples, it can apply to the purchase of used industrial equipment, commodities, or freight capacity. Virtual enterprises are business arrangements in which trading partners separated by geography and expertise are able to engage in complex joint business activities, as if they were a single enterprise. One example would be true supply chain integration, where planing and forecast data are transmitted quickly and accurately throughout a multi-tier supply chain. Another example is a large automotive OEM doing designing cars by doing concurrent engineering projects with the suppliers. These methods of doing business differ from traditional commerce because electronic commerce combines information and telecommunications technology with business processes to make it possible to do business in ways that could not otherwise be done. As an example: suppose a company wishes to set up a Website through which customers can order products. Technological requirements for this system would include:
Without the accompanying business processes this technology is, at best, useless. At worst the technology can be destructive if it bogs down operations and alienates customers. To make the system work a business process is needed to assure that the information on the site matches what is actually happening within the company’s order processing operations. So at a minimum the site needs to be continually updated so that customers can trust what they read about products -- specifications, availability, and prices. What are the technologies of EC? While many technologies can fit within the definition of "electronic commerce," the most important are electronic data interchange (EDI), bar codes, electronic mail, the Internet, the World Wide Web, product data exchange, and electronic forms. Other technologies, such as IRC, video conferencing, white boarding and internet based meeting software, are emerging. Although some of them show some interesting promise, they are not yet widespread and are beyond the scope of this Implementation Guide. Electronic Data Interchange (EDI) is the computer-to-computer exchange of structured business information in a standard electronic format. Information stored on one computer is translated by software programs into standard EDI format for transmission to one or more trading partners. The trading partners’ computers, in turn, translate the information using software programs into a form they can understand. Bar Codes are used for automatic product identification by a computer. They are a rectangular pattern of lines of varying widths and spaces. Specific characters (e.g. numbers 0-9) are assigned unique patterns, thus creating a "font" which computers can recognize based on light reflected from a laser. The most obvious example of bar codes is on consumer products such as packaged foods. These codes allow the products to be scanned at the check out counter. As the product is identified the price is entered in the cash register, while internal systems such as inventory and accounting are automatically updated. The special value of a bar code is that objects can be identified at any point where a stationary or hand held laser scanner can be employed. Thus the technology carries tremendous potential to improve any process requiring tight control over material flow. Good examples are shipping, inventory management, and work flow in discrete parts manufacturing. Electronic Mail allows messages composed by an individual to be sent in digital form to other recipients via the Internet. The Internet is a decentralized global network of millions of diverse computers and computer networks. These networks can all "talk" to each other because they have agreed to use a common communications protocol called TCP/IP. The Internet is a tool for communications between people and businesses. The network is growing very, very fast and as more and more people are gaining access to the Internet, it is becoming more and more useful. The World Wide Web is a collection of documents written and encoded with the Hypertext Markup Language (HTML). With the aid of a relatively small piece of software (called a "browser"), a user can ask for these documents and display them on the user’s local computer, even if the document is on someone else’s computer across the ocean. HTML documents (or "pages," as they are called) can contain many different kinds of information such as text, pictures, video, sound, and pointers which take users immediately to other Web pages. Because Web pages are continually available through the Internet, these pointers may call up pages from anywhere in the world. It is this ability to jump from site to site that gave rise to the term "World Wide Web." Product Data Exchange refers to any data that is needed to describe a product. Sometimes that data is in graphical form, as in the case of pictures, drawings and CAD files. In other cases the data may be character based (numbers and letters), as in the case of specifications, bills of material, manufacturing instructions, engineering change notices and test results. Product data exchange differs from other types of business communications in two important ways. First, because graphics are involved users must contend with large computer files and with problems of compatibility between software applications. (The difficulty of exchanging CAD files from one system to another is legendary.) Second, version control very quickly gets very complicated. Product designs, even late in thedevelopment cycle, are subject to a great deal of change, and because manufacturing processes are involved, even small product changes can have major consequences for getting a product into production. Electronic Forms is a technology that combines the familiarity of paper forms with the power of storing information in digital form. Imagine an ordinary paper form, a piece of paper with lines, boxes, check-off lists, and places for signatures. To the user an electronic form is simply a digital analogue of such a paper form, an image which looks like a form but which appears on a computer screen and is filled out via mouse and keyboard. Behind the screen, however, lie numerous functions that paper and pencil cannot provide. Those extra functions come about because the data from electronic forms are captured in digital form, thus allowing storage in data bases, automatic information routing, and integration into other applications. As an example, a supplies form may be filled out by the requester and automatically sent to a supervisor for approval. Once approved, the actual order may be input into an EDI translator, and go to the vendor by means of a structured X12 EDI transaction. What are the business opportunities offered by EC technologies? EC is being implemented because it is an excellent method of dealing with forces that are increasingly presenting SMEs with new problems and new opportunities. Global sourcing and the global marketplace, combined with the communication potential of the Internet, are presenting business with wider markets and more competition. The intertwined trends of outsourcing, fast-response manufacturing and supply chain integration will increase demands for collaboration and coordination technologies. Governments are moving toward electronic reporting for payroll, compliance and regulation. Competitive pressures will increase manufacturers’ needs for data integration and coordination-enhancing technologies. Information will become an ever more important competitive asset, much to the advantage of companies who can manage, manipulate, and mine their data. As businesses implement EC to accommodate to these trends, applications of EC will be seen to fall into nine categories, as shown by the examples in the following matrix. (Numbers in cells are keyed to the immediately following explanatory text.)
EC business process, product/process x business impact 1. X-12 and EDIFACT based EDI is used to send traditional purchase orders faster and with less expense. Without EDI purchase orders would still flow, but not as quickly and at greater expense. 2. Your company uses the Web to search for commodity items and used equipment that are listed in on-line catalogues. This makes shopping easier and faster, but paper catalogues are available and more than adequate for most tasks. 3. Engineering change notices, complete with graphics, can be sent via mail or over a network. As with purchase orders the information would arrive anyway, but is transmitted more quickly (and perhaps in a more usable form) through a network. 4. Evaluated receipts settlement, a system that pays on the bases of "evaluating" documents such as purchase orders and shipping receipts, and thus does away with a formal invoicing process. 5. The introduction of large-scale just-in-time production systems. 6. Concurrent engineering among your design teams and your customers’. 7. Vendor-managed inventory based on point of sale data capture. 8. Groups of non-competing suppliers form "one stop shopping" systems for mutual customers.
Why is it important to simultaneously manage technical and organizational change? New technology has implications for how organizations work. Sometimes the implications are minor and work themselves out with minimal planning. Sometimes deliberate effort is needed to assure a smooth transition to technology that serves an organization well. Early in the implementation process companies would do well to develop a general sense of how much effort will be needed to manage the organizational consequences of the change. To illustrate let’s consider how the World Wide Web might be used by Wiffle Enterprises, an SME that specializes in fabricating gas and water tanks for the automotive industry. Most of their work is outside the mainstream of automobile manufacturing, they work for makers of specialty trucks (e.g. heavy equipment carriers) and motorcycles. They also do some work for the snowmobile industry. Wiffle does some custom work that requires engineering, but about 60% of its business involves items they have built before. It’s a short production run job shop type operation. Wiffle is considering two options for its Web use. Web Options for Wiffle Enterprises Option 1: Static Web Pages In this situation Wiffle’s objective is to advertise itself on the Web in order to make itself known to potential customers. For this purpose the Web site need contain information that is fairly stable - general description of product line and markets, likely lead times, location, brief history, facilities, and location and contact information. While all Web pages need provision to be updated, the maintenance burden here is quite light. More important, the existence of the Website has few implications for other parts of the company. Sales, Marketing, Shop Floor Scheduling, Order Processing, Information Systems, Accounting -- all can go on as they did before, with hopefully a little more work because the Website brought in some new business. Option 2: Website for Electronic Commerce Now that Wiffle has become comfortable with having a Website, it is beginning to want more. The new plan is to identify customers with whom Wiffle a trustworthy relationship, and allow them to place orders through the Web site. While Wiffle can’t provide this service for all its products, it can do so for stock items that require no engineering or customization. For the stock items customers can use the Website to check on availability, delivery time, and prices (complete with price breaks for quantities). Because the product line is so diverse (lots of tank designs are in use), Wiffle plans to limit its on-line catalogue to relatively few items that account for fifty percent of its sales. As an additional service to its customers, Wiffle also plans to allow customers to check on order status via an Email dialogue box. There is commercial off-the-shelf technology that Wiffle can use to set up this Website. The technology is not that complex and its not that expensive. Option two is certainly appealing, but consider the business process needed to make it work. Maintaining the site: Because the information is so variable, and because accurate information is so important, Wiffle will need to maintain its site carefully. The company will have to decide whether to maintain this capacity in-house, or to outsource it. If the in-house path is taken a host of follow-on decisions are necessary. Does Wiffle have qualified people to do the work, and if so what provision should to be made for the work they used to do? If people are not available a hiring process must be started. Whether existing personnel or new hires are used, it is quite likely that training will be needed in whatever electronic commerce system is implemented. Then there is the question of reporting relationships. Working on the Website seems like a job to be overseen by the IS department, but the content involved is within the purview of Sales and Marketing, so who should be in charge? There is no absolute right or wrong answer, and a decision will have to be made that makes the most sense for Wiffle’s particular circumstances. Business arrangements with customers: Wiffle is certainly not willing to let anyone order anything through its Web site. This is a service offered to customers with whom there is an established relationship, and then, only for certain items. The company will have to decide what the criteria are, and which customers qualify. It will also have to make sure those customers understand their options, and come to an agreement on how the system will work. Will orders that come through the Web have be followed up in writing within a particular period of time? If so who is in charge of reconciling the electronic and hard copy orders? Are there dollar limits? How will customers be made aware of their new ordering option? Other than convenience, and skill should Wiffle offer other inducements for using the Web system? None of these issues may be complicated, but all require deliberate action, and each is critical to the success of the system. Internal process changes: Consider some of the changes that are likely. Order Processing will have to adjust to an additional and entirely new way for orders to enter the company. The role and influence of the sales force may change. Account management will have to provide the fast response to customer requests that people come to be expected when they use email. If Wiffle is like many small companies its ability to track inventory is unreliable and will need considerable improvement before information on inventory can be made easily accessible to the public. Site maintenance, business arrangements and internal process change -- taken together their cost in dollars and effort far exceed any investment in technology. In its totality the move to Web-based sales represents a serious business decision that Wiffle industries must consider in terms of return on investment, opportunity costs, employee moral, and risk. These decisions have little to do with technology, and everything to do with business objectives, company structure, the work people do, and how employees come by the knowledge they need. The Jackson Window Case Jackson’s management asked its local MEP center to help them get a general sense of what various EC technologies could to do help the company better serve its customers. Going on instinct alone, the field agent realized that the most important technologies for Jackson would probably be EDI, the Web and email. All three will help Jackson communicate with customers and suppliers. In addition, email will help solve problems with internal coordination. To help educate Jackson about these technologies she recommended one training session on each. These she found by checking the centralized list of courses offered by Electronic Commerce Resource Centers, and through a call to a local community college known for its active efforts in Internet training. Finally, she explained to Jackson that if they were going to make good use of the Internet and email, they would need to install a local area network. The next step was to do a quick business analysis of Jackson’s situation, asking about how their business was changing, what their major worries were, and what opportunities they saw over the next few years. Based on this analysis the field agent made suggestions about five EC technologies. None of these suggestions could count as specific recommendations because each would be a major effort in its own right, and a more rigorous analysis of business need and implementation feasibility would be required to make specific choices. Still, it was useful to point Jackson’s management in the right direction and to start them thinking about specific choices. To summarize her suggestions, she produced the following table.
Business Analysis OutlineIntroduction The first section of this chapter, "identifying targets of opportunity" is identical to its counterpart section in the Primer. The following section ("Determining Feasibility") builds on the Primer in two ways. First, it drives the discussion of technological alternatives to a finer level of detail. Second, it puts additional emphasis on "organizational feasibility" of proposed EC changes . Note that this section, which is titled "Business Analysis", contains a discussion of technology, a topic that seems to belong later, in the section on "Requirements Analysis". This is a prime example of feedback loops and iterations between stages of the EC implementation. While rigorous requirements definition is truly a topic in its own right, it is also the case that developing a realistic business plan for EC requires some input from a requirements analysis exercise. Identifying Targets of Opportunity The question of how to identify a target of opportunity for electronic commerce is as broad and difficult to answer as any general question about making a business more competitive. How for instance would you answer the familiar question: How can I generate more business? Brainstorming and inspiration are useful methods for answering this question, but the best solution will come from combining these methods with a systematic approach. As an example it would make sense to try to complete tables 1 and 2.
The more unfamiliar the territory the greater the importance of system over inspiration. Thus to help answer the question about using EC to benefit your business we offer a systematic approach based on our virtual focus group on trends in electronic commerce. The key to discovering how EC can help your business is to answer two questions.
A useful way to answer these questions is to place your list of possible changes within the table 3. (The previous chapter "Awareness" has examples of business processes that can populate this matrix.)
We recommend an approach that oscillates between the systematic and the inspirational. As ideas are generated they can be placed within the table, thus providing a sense of value and risk. Working from the table, it becomes possible to decide on an acceptable level of benefit and risk, and to then search your business environment for possibilities that fit your comfort level. Determining the feasibility of alternative EC plans As business needs are determined it is necessary to establish the technological feasibility of various EC plans which could meet those needs. The starting point should be a clear sense of what functions each EC technology can provide to improve business functioning. We summarize these in Table 4.
Once a company has a sense of business objectives and technological possibilities, the two can be honed against each other and fashioned into a set of priorities. (A flow chart of the logic of this process appears in Chapter 6 of the Primer.) As an aid in performing this analysis, Table 5 will be useful.
Note that assessments made in Table 2 are "first impressions", i.e. they are meant to convey a general sense of possibilities but not to bog the process down in detail. The idea is to give people the confidence to proceed with the development of an overall plan. Of course later analysis may force a revision of these first impressions, but one is much better off taking the chance that revision will be necessary than to work out all the details at each step. Its almost a certainty that too quick a descent into detail will lead to a loss of direction, missed opportunities to reach consensus on important business issues, and poor coordination between business change and technological change. (Early in the next chapter we will present an example of how a look at detail can force a revision in the business analysis.) The Jackson Window Case Not surprisingly, Jackson’s situation (as first described in Chapter 1) has been the subject of much discussion within the company, and is well known to management. With respect to customers the problems lie within the last three rows of table 1 and the first row of table 2.
The system manager at Jackson Windows decided that some fundamental decisions would need to be made about what kind of system Jackson would build for itself. After consulting with the President, he pulled together a small steering committee of upper managers including the President and owner of the company to help him make these decisions. Few decisions at Jackson could be made without the consent of the President anyway, so it made sense to include him in the group to begin with. The group quickly decided to construct a Local Area Network to support both Electronic Commerce applications and the new Manufacturing System that Jackson would implement. This decision (and many to follow) were made with a reasonable sense of cost, but without benefit of a rigorous analysis of costs and benefits. Such an analysis would be superfluous because many of the choices reflect management’s pre-existing commitment to change the company. The fundamental decision had been made based on a business judgement about where the market was going. Also, in many respects the option of standing pat did not exist. For example, the Manufacturing System at Jackson had to be replaced because it was old, not Year-2000 capable and no longer supported because the company that wrote the software had gone out of business. The group then proceeded to filter the MEP field agent’s recommendations (from chapter 2) through table 4 (basic technological choices), and table 5 (feasibility assessment.) Their conclusions follow. EDI The group decided that it wanted to do EDI with its EDI-capable customers but that the volume of these transactions did not justify a fully integrated system. Also Jackson's internal systems were not in good enough shape yet to let the information flow directly from the customers into Jackson's Order Entry system. So to start with, Jackson decided to do rip and read EDI. The steering committee also decided to concentrate on the Material transaction sets, although the President did decide to ask the controller to investigate doing some financial EDI transactions. The long term plan was to integrate EDI and to do EDI with Jackson's suppliers. Developing a schedule to implement this plan must be determined by progress in implementing the new manufacturing system.
World Wide Web The steering committee debated long and hard about how Jackson should develop its Web presence and came up with the following conclusions: Immediate action: Near future: As soon as this Web presence was put in place Jackson would start to put its catalogue on the Web, a plan which requires that the ever-changing catalogue be updated regularly. Accurate updates will be available because computerizing the catalogue in an easily accessible database was already a priority project for the marketing department.
Long term: In the long run the company wants to allow customers to order on-line and to obtain order information (like anticipated ship-date or shipment history). But Jackson must first become a much more disciplined company with much more efficient and computer oriented business processes before this can happen. So the steering committee decided that Jackson will go from a Passive Web Site, with lean content immediately to a Passive Web Site with rich content in the intermediate term to an Active Web Site with rich content in the long term. Barcodes The steering committee decided that most of the implementation of any bar coded systems would need to be driven by the implementation of the new Manufacturing System. Email. There was little debate about the need for an effective communication tool, but there was disagreement about who should have access to the new email system. What was debated was who was to have access to email in the company. The group concluded that immediate access should be given to all office, plant supervision, and shipping/receiving personnel. Some members of the committee were inclined toward extending email to all employees (including hourlies), but the group decided to suspend judgement until they could observe the consequences of initial implementation. Product Data Exchange. The steering committee decided that email attachments were an ideal method for exchanging files with its customers and suppliers. Their reasoning was that email was being installed anyway and that people were comfortable with that technology. Also, because the number and size of CAD files requiring transfer would be relatively small, there was no need for FTP or other less familiar methods of data transmission. The group decided to postpone making any other decisions about issues involved with PDX (such as the type of files to be exchanged, their formats etc). They felt that the issue was not critical even though they knew that automated transfer would increase efficiency and contribute to some lead time reduction. Also, dealing with PDX would be a time consuming problem that would move the groups’ attention away from its other priorities. So rather than set policy, they decided to leave decisions to personal negotiations between individual engineers and their customer contacts. Electronic Forms. The group decided to postpone the implementation of an e-forms system. This was done not because the technology didn't promise to deliver some efficiencies, but to preserve the available resources, especially in terms of people. Jackson Windows is implementing a new network, an Email system and EDI, all at the same time as its engineering people are changing the design of product line to support the change in manufacturing from build-to-Stock to Build-to-Order. While all this is going on, the manufacturing people are changing the plant to support this same change. Finally the whole company is in the process of implementing a new enterprise wide manufacturing package. There is only so much change that any organization can manage at once.
Requirements Definition Outline
Bridging Business Priorities and Implementation Detail The previous chapter brought you to basic decisions about what kind of EC will serve your company best. The next chapter will provide nitty gritty detail on implementation. This chapter deals with a level of abstraction that fills the gap between statements about business need and specific instructions for implementation. Thus, we are advocating a three level approach to implementation.
Why use a layered approach? To begin the EC process decision makers need an understanding of what EC functionality will serve their company best, and the confidence needed to commit to a serious exploration of feasibility. Too quick a descent into detail will lead to a loss of direction, missed opportunities to reach consensus on important business issues, and poor coordination between business and technological change. Later analysis may force a revision of initial decisions. Ultimate success, however, is much more likely by taking the chance that revision will be necessary than by risking the loss of direction that can come from too much detail too early in the process. Of course in the real world of business decision making there cannot be a clean, linear progression through set stages. The world is too complicated and too fast moving. Too much happens that is beyond our control. Too many constraints are set before us. We may have easy access to an expert in a useful technology but no expert in a more desired technology. There may be pressure to justify a system or a business process that has already been implemented. There may be budget available for one set of activities but not for another. For these and for many, many other reasons we are usually forced into starting a decision making process at many points far away from its logical beginning. What is important is not to start at the beginning and proceed in lockstep, but to realize that there is a beginning and there is a process. With that knowledge it becomes possible to understand where the opportunities are for systematic planning, to take calculated risks, and to maximize the possibilities for a worthy outcome. As an illustration of imposing a process when forced to start in the middle, consider the case of a company faced with the following:
If the company were to start with a business analysis exercise, it would probably not choose EDI as its first priority. But here it is, faced with a reality pushing it toward EDI. The company finds itself in the middle of "requirements definition", considering two viable options: 1- A bare minimum rip and read EDI system, and 2- an EDI service bureau. Option two can be quickly dismissed because of the high transaction volume. The cheapest and fastest way to satisfy the customer is to go with rip and read. Suppose, however, that the company backed up and asked itself a question. "What would things look like if we looked at our business needs for EC?" They might conclude that what is really needed is a Web presence to help advertise, communicate with customers, and find suppliers. That realization would lead to holding off on EDI for the time needed to set up a Web site. Then an Internet based EDI system could be implemented which would not only be relatively inexpensive, but would also help justify the investment in the networking infrastructure. The desirable solution did not come because the company started at the beginning. It came because they realized that there was a beginning, and was able to use that knowledge it its planning efforts. Interactions Between "Business Analysis" and "Requirements Definition" Successful requirements definition requires careful attention to the strong interactions that it has with business analysis, as illustrated in the following example. Example: Interaction of "business analysis" and "requirements definition" As a result of a business analysis exercise a company decides that it is in their interest to set up an information-rich Web site that provides customers and potential customers with information about product availability. They decided this plan was feasible by working with the "feasibility assessment matrix" in the previous chapter. The company reasoned that the plan was feasible because they already had a Web site in operation, a trustworthy expert to maintain the site, and a data base on product availability. Upon more careful analysis, however, a revision was required. The revision process is illustrated in the "feasibility matrix" presented in table 1. (For illustrative purposes we only present a partial matrix in this example.)
A Process for Requirements Definition The first step is to identify the necessary technology and business process. A convenient way to do this is to think of a relationship matrix as depicted in Table 2.
While it is not necessary to split needs by "EC", "support" and "business", it is enlightening to do so. By separating specialty and generic technology, one gets a sense of where technology can be put to multiple purposes. The reason to include "business process" is to assure that the total implementation effort is planned and budgeted. Out of the matrix should come a list of specifics, e.g. EC technologies: EDI translator that can run on an AS400 platform, ability to accept credit card payment on the Web, individual email accounts for all our employees. Supporting technologies: MRP system to accept incoming EDI data, database on product availability to support Web-based on-line purchasing system. Business process: Evaluated receipts settlement to eliminate the invoicing process, agreement with trading partners to exchange CAD files in digital form via email attachments. What level of detail is needed? While there is no specific answer, a useful rule of thumb is: enough detail to get reasonable estimates of cost, level of effort, and time frame. The next step is to actually determine what resources are needed (and to revise requirements if need be.) To do this we need to track requirements against resources, as depicted in table 3.
All resources can be assessed in terms of dollars, and should be. Most (except for the grayed out cells) can also be assessed in terms of hours for particular kinds of people. It is useful to make both assessments because while dollars can buy time, in reality the two might not be interchangeable. For example designing a Web based catalogue may require the expert knowledge of one or two key staff in Purchasing, and those people may not have the time to devote to the project. No matter how many dollars were available, getting that expertise in a reasonable period of time might be impossible. Make or Buy? For EC technology the only sensible option is (almost always) to buy commercial off the shelf products. EC support services, however, can often be subject to a make/buy analysis. Examples include Web page maintenance, integration "glue" for in-house applications, process engineering expertise, and training. In cases like these a company may well have the expertise and dollars to make a choice between doing the work themselves or buying it. How to choose? One obvious criteria is money, i.e. how much will it cost to implement and maintain the system if it were done in-house or by outsiders? Money, however, is not the only consideration. Other important factors include: Business Impact
Quality of Service
Jackson Window Example The Systems Manager at Jackson now went to work detailing the cost involved in the decisions described in the previous chapter. He had to do much research and was often seen on the World Wide Web reviewing comparisons of software or hardware alternatives. Because bar codes and electronic forms were not set as a high priority, he did not concern himself with these technologies. As part of the implementation of the Manufacturing System, the decisions involved in choosing the hardware for a Local Area Network had already been made. The publisher of the Manufacturing Software dictated some of these choices. Jackson chose to run on its network on a Windows NT Server with workstations using Windows 95/98. EDI As a result of business analysis and feasibility assessment (chapter 3) the plan was:
Cast as a requirement that can be entered into table 2, the language translates into a solution with initial low cost that is combined with the capacity to easily scale up in terms of transaction volume, transaction variety, and integration. The systems manager decided to choose General Electric as the VAN (Value Added Network) supplier and make use of the GEWeb product. This product translates the customer's EDI transactions sets (in this case Purchase Orders) into human readable Web pages. The user then prints the Web page and enters the data into his own Order Entry system. After shipment of the order, the user can key the Advance Shipment Notice data into a Web based form, which then gets translated into EDI and send to the customer. This satisfied the customer's requirement for now, and will cost less then $100/Month. Later, when Jackson is ready to integrate EDI data automatically into the Order Entry system, they would acquire the in-house capability to generate and receive X-12 based transaction sets. World Wide Web As a result of business analysis and feasibility assessment (chapter 3) the plan was:
The feasibility analysis did not look good for in-house administration of the site.
Given these facts, the system manager decided to let another company host the Jackson Web site. The Web hosting service will work with Jackson's existing advertising agency and marketing department to create the content for the site. The system manager carefully checked out the Web hosting company, its pricing, its reputation and its capabilities and then decided that this was the best solution for Jackson. The immediately result of this decision was a reduced bandwidth requirements for his Internet connection to the LAN. This significantly off-set the cost of doing business with the Web-hosting company. Without Jackson's Web traffic, the only traffic into and out of the LAN will be WWW browser traffic from LAN users and e-mail. Thus a simple 128K ISDN connection will probably be enough. Electronic Mail Expected use in the short and long term are similar. In both cases there will be many users, thus necessitating a separate account for each user. Ideally the system would be easy to implement, easy to administer, and easy to use. It soon became clear that all three of these requirements could not be met, and the Systems Manager opted for emphasizing ease of use. After much consideration he chose MS Exchange Server, since it came with the MS Back Office application that he purchased with his NT server software. Because it was bundled with Back Office, cost at least, was low. Not so ease of implementation. The main problem with this system was that MSExchange is hard to set up, so the system manager decided to hire an MS Exchange consultant. Although $1200/day was considered expensive by Jackson management, it was the most sensible solution for this problem. The Manager reasoned that by working with the consultant (and getting a little extra training) his staff would learn enough about administration to manage the task. He knew that system administration would take up a reasonable percentage of someone’s time, but that would probably not be much higher than for other systems which could be chosen. Although maintaining group calendars was never a high priority at Jackson, the system manager assumed that this might become a popular feature with the users over time. Earlier, the company had decided that every 'MS Office-equipped' desktop computer was to have the MS Office suite of software. As a result of this, all workstations were equipped with Microsoft Outlook, an excellent client designed for MSExchange. So there was little out of pocket software expenses. | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||