The following are techniques used to determine systems requirements and a hypothetical situation in which each would be the best method. In a real-life situation, it may be useful, however, to use several of these in some combination, with only one being the principal method.

Documents: When hard-copied forms pass much of the available information, examination of the documents will provide insight into an organization’s structure and data requirements. For example, a state run bureaucracy may have many forms. Upon examination, it may be determined that much of the information on each form is redundant (such as a project name, a project number, an engineering number, and a project year – when a good database should be able to reference a project based on a single index number). Other portions of the form may be continuously left blank. The obsolete data does not need to be part of a new system. The State of Kentucky’s SAS-5 form for example has a space for “agency number”, and a space for “organization number”. The AI, the PBU, and the Project are also redundant. One project number should be used, and the other necessary information should be kept on a separated database. The project description should be kept to insure that the project number is correct.

Interview: In an instance where getting opinions about a system is important, an interview would be the most powerful tool. For example, in an expanding firm, getting opinions about which direction the company is headed is critical to designing a new system or adding improvement to the old. Consider a retailer setting up a new web site. When Wal-Mart finally went on line in the late ‘90’s, they faces many issues. Will they want online ordering and billing? Will they want to suppliers to view their company’s inventory so it can be automatically reordered? What specific information will the management need? What question is to be addressed by the new system? The manager overseeing the operation should be interviewed. She may want a site for customers to order from, but she does not feel that she wants more flexibility in selecting suppliers, and does not want to grant them access to her inventory information. Information in an interview can flow both ways, and she may not even be aware that the orders from the website could be tied into her internal inventory, to automatically show out of stocks to potential customers, and to update her records for financial statements. She may have lofty expectations of earnings from a website, and it may have to be pointed out that spending on online advertising will have to be a part of a website strategy, and some online customers will be erosion of shoppers in her brick & mortar store.

Joint Application Design (JAD): This method would work well with a group of professionals, with technically sophisticated data needs. Examples include a biotech firm or a university. These knowledge-based organizations are more amenable to problem solving ideas coming from the users. Also, since the users have specialized data needs that may not be readily known to central management, the users themselves would have a strong demand to have input into the creation of a new system. In the case of Pfizer, for example, a researcher studying Benadryl may want to see the actual “hard” data of a study, and the others in the group may decide that only the aggregate statistics should be shown, to reduce the degree to which the results of individual test subjects can be tampered with. For instance, if the entire database was open, a researcher may decide to reject those subjects that have a family history of allergies. A decision on such a point will have to be reached before the system can be developed. Can security features be put into place? Can the group work with just the statistics? The technical leaders from the team can have an opportunity to discuss with each other what the objectives for the system are. The system designer will probably not be familiar with the relevant statistics, and should discuss with the teams which ones they use more frequently. They can agree on which exactly which statistics had meaning and relevance to their studies. The department heads may wish to see more financial information for budgetary issues as well. Financials may not be foremost on the mind of a department head until she hears it brought up by another department head first. A JAD can be more productive than many separate interviews because of the team thought process.

Questionnaire: If there are a large number of people from which to solicit information (too large for a JAD), a questionnaire may be the answer. If the information you seek is customer information, using Wal-Mart for example again, other methods may not work as easily. Online customers may be asked to fill out a short survey to receive a small discount on their purchase. Information may be gathered to ascertain:

· Why they were shopping online vs. at a brick & mortar?

· Why they were at the client web site vs. a competitor?

· How difficult it is to navigate the web site?

· How did they learn about the website, from TV, print ads, word-of-mouth, through a search engine or online banner link?

Direct Observation: When it is important to determine the attitude of the company for a new or improved system, direct observation may be the key. On the issue of security, a laid-back, jeans-and-T-shirt firm may not want or need password protection and limited screen access for certain employees. On the other hand, a firm that projects authority and has a formal, top-down power structure, will need to be handled in a way that is sensitive to the demands of the manager, with less input from the end user. A firm that gives off the impression of being resistant to change will need a system that closely resembles their old one. A law firm, for example, with a few top partners and several secretaries and paralegals may have those qualities.