Transaction Processing Systems

A firm’s integrated information system starts with its transaction processing system (TPS). The TPS receives raw data from internal and external sources and prepares these data for storage in a database similar to a microcomputer database but vastly larger. In fact, all the company’s key data are stored in a single huge database that becomes the company’s central information resource. As noted earlier, the database management system tracks the data and allows users to query the database for the information they need.

Exhibit 13.6 A Company’s Integrated Information System (Attribution: Copyright Rice University, OpenStax, under CC BY 4.0 license.)

The database can be updated in two ways: batch processing, where data are collected over some time period and processed together, and online, or real-time, processing, which processes data as they become available. Batch processing uses computer resources very efficiently and is well-suited to applications such as payroll processing that require periodic rather than continuous processing. Online processing keeps the company’s data current. When you make an airline reservation, the information is entered into the airline’s information system, and you quickly receive confirmation, typically through an e-mail. Online processing is more expensive than batch processing, so companies must weigh the cost versus the benefit. For example, a factory that operates around the clock may use real-time processing for inventory and other time-sensitive requirements but process accounting data in batches overnight.

Decisions, Decisions: Management Support Systems

Transaction processing systems automate routine and tedious back-office processes such as accounting, order processing, and financial reporting. They reduce clerical expenses and provide basic operational information quickly. Management support systems (MSS) use the internal master database to perform high-level analyses that help managers make better decisions.

Information technologies such as data warehousing are part of more advanced MSSs. A data warehouse combines many databases across the whole company into one central database that supports management decision-making. With a data warehouse, managers can easily access and share data across the enterprise to get a broad overview rather than just isolated segments of information. Data warehouses include software to extract data from operational databases, maintain the data in the warehouse, and provide data to users. They can analyze data much faster than transaction-processing systems. Data warehouses may contain many data marts, special subsets of a data warehouse that each deal with a single area of data. Data marts are organized for quick analysis.

Companies use data warehouses to gather, secure, and analyze data for many purposes, including customer relationship management systems, fraud detection, product-line analysis, and corporate asset management. Retailers might wish to identify customer demographic characteristics and shopping patterns to improve direct-mailing responses. Banks can more easily spot credit-card fraud, as well as analyze customer usage patterns.

According to Forrester Research, about 60 percent of companies with $1 billion or more in revenues use data warehouses as a management tool. Union Pacific (UP), a $19 billion railroad, turned to data warehouse technology to streamline its business operations. By consolidating multiple separate systems, UP achieved a unified supply-chain system that also enhanced its customer service. “Before our data warehouse came into being we had stovepipe systems,” says Roger Bresnahan, principal engineer. “None of them talked to each other. . . . We couldn’t get a whole picture of the railroad.”

UP’s data warehouse system took many years and the involvement of 26 departments to create. The results were well worth the effort: UP can now make more accurate forecasts, identify the best traffic routes, and determine the most profitable market segments. The ability to predict seasonal patterns and manage fuel costs more closely has saved UP millions of dollars by optimizing locomotive and other asset utilization and through more efficient crew management. In just three years, Bresnahan reports, the data warehouse system had paid for itself.^[1]

At the first level of an MSS is an information-reporting system, which uses summary data collected by the TPS to produce both regularly scheduled and special reports. The level of detail would depend on the user. A company’s payroll personnel might get a weekly payroll report showing how each employee’s paycheck was determined. Higher-level mangers might receive a payroll summary report that shows total labor cost and overtime by department and a comparison of current labor costs with those in the prior year. Exception reports show cases that fail to meet some standard. An accounts receivable exception report that lists all customers with overdue accounts would help collection personnel focus their work. Special reports are generated only when a manager requests them; for example, a report showing sales by region and type of customer can highlight reasons for a sales decline.

Decision Support Systems

A decision support system (DSS) helps managers make decisions using interactive computer models that describe real-world processes. The DSS also uses data from the internal database but looks for specific data that relate to the problems at hand. It is a tool for answering “what if” questions about what would happen if the manager made certain changes. In simple cases, a manager can create a spreadsheet and try changing some of the numbers. For instance, a manager could create a spreadsheet to show the amount of overtime required if the number of workers increases or decreases. With models, the manager enters into the computer the values that describe a particular situation, and the program computes the results. Marketing executives at a furniture company could run DSS models that use sales data and demographic assumptions to develop forecasts of the types of furniture that would appeal to the fastest-growing population groups.

Companies can use a predictive analytics program to improve their inventory management system and use big data to target customer segments for new products and line extensions.

Exhibit 13.7 Decision support systems help businesses by providing quantitative data and predictive models that aid problem-solving and decision-making. Now the health-care industry wants this technology in hospitals to improve the practice of medicine. Spearheading the effort for a clinical decision-support system is the American Medical Informatics Association, which believes a national DSS could help physicians with diagnosing and treating illnesses. What are the pros and cons to having medical professionals rely on a DSS for help in treating patients? (Credit: Axelle Geelen/ flickr/ Attribution 2.0 Generic (CC BY 2.0))

Executive Information Systems

Although similar to a DSS, an executive information system (EIS) is customized for an individual executive. These systems provide specific information for strategic decisions. For example, a CEO’s EIS may include special spreadsheets that present financial data comparing the company to its principal competitors and graphs showing current economic and industry trends.

Expert Systems

An expert system gives managers advice similar to what they would get from a human consultant. Artificial intelligence enables computers to reason and learn to solve problems in much the same way humans do, using what-if reasoning. Although they are expensive and difficult to create, expert systems are finding their way into more companies as more applications are found. Lower-end expert systems can even run on mobile devices. Top-of-the-line systems help airlines appropriately deploy aircraft and crews, critical to the carriers’ efficient operations. The cost of hiring enough people to do these ongoing analytical tasks would be prohibitively expensive. Expert systems have also been used to help explore for oil, schedule employee work shifts, and diagnose illnesses. Some expert systems take the place of human experts, whereas others assist them.