There are three basic decision models used in gdss: operational, strategic, and financial.

Chapter 12 Enhancing Decision Making

LEARNING OBJECTIVES

After reading this chapter, you will be able to answer the following questions:

1. What are the different types of decisions and how does the decision-making process work?

2. How do information systems support the activities of managers and management decision making?

3. How do decision-support systems (DSS) differ from MIS and how do they provide value to the business?

4. How do executive support systems (ESS) help senior managers make better decisions?

5. What is the role of information systems in helping people working in a group make decisions more efficiently?

CHAPTER OUTLINE

12.1 DECISION MAKING AND INFORMATION SYSTEMS

Business Value of Improved Decision Making

Types of Decisions

The Decision-Making Process

Managers and Decision Making in the Real World

12.2 SYSTEMS FOR DECISION SUPPORT

Management Information Systems (MIS)

Decision-Support Systems (DSS)

Data Visualization and Geographic Information Systems

Web-Based Customer Decision-Support Systems

Group Decision-Support Systems (GDSS)

12.3 EXECUTIVE SUPPORT SYSTEMS (ESS) AND THE BALANCED SCORECARD FRAMEWORK

The Role of Executive Support Systems in the Firm

Business Value of Executive Support Systems

12.4 HANDS-ON MIS PROJECTS

Management Decision Problems

Improving Decision Making: Using Pivot Tables to Analyze Sales Data

Improving Decision Making: Using a Web-Based DSS for Retirement Planning

LEARNING TRACK MODULE

Building and Using Pivot Tables

Interactive Sessions:

Too Many Bumped Fliers: Why?

Business Intelligence Turns Dick’s Sporting Goods into a Winner

EASTERN MOUNTAIN SPORTS FORGES A TRAIL TO BETTER DECISIONS

Founded in 1967 by two rock climbers, Eastern Mountain Sports (EMS) has grown into one of the leading outdoor specialty retailers in the United States, with more than 80 retail stores in 16 states, a seasonal catalogue, and a growing online presence. EMS designs and offers a wide variety of gear and clothing for outdoor enthusiasts.

Until recently, however, the company’s information systems for management reporting were dated and clumsy. It was very difficult for senior management to have a picture of customer purchasing patterns and company operations because data were stored in disparate sources: legacy merchandising systems, financial systems, and point-of-sale devices. Employees crafted most of the reports by hand, wasting valuable people resources on producing information rather than analyzing it.

After evaluating several leading business intelligence products, EMS selected WebFOCUS and iWay middleware from Information Builders Inc. EMS believed WebFOCUS was better than other tools in combining data from various sources and presenting the results in a user-friendly view. It is Web-based and easy to implement, taking EMS only 90 days to be up and running.

IWay extracts point-of-sale data from EMS’s legacy enterprise system running on an IBM AS/400 midrange computer and loads them into a data mart running Microsoft’s SQL Server database management system. WebFOCUS then creates a series of executive dashboards accessible through Web browsers, which provide a common view of the data to more than 200 users at headquarters and retail stores.

The dashboards provide a high-level view of key performance indicators such as sales, inventory, and margin levels, but enable users to drill down for more detail on specific transactions. Managers for merchandising monitor inventory levels and the rate that items turn over. E-commerce managers monitor hour-by-hour Web sales, visitors, and conversion rates. A color-coded system of red, yellow, and green alerts indicates metrics that are over, under, or at plan.

EMS is adding wikis and blogs to enable managers and employees to share tips and initiate dialogues about key pieces of data. For example, in identifying top-selling items and stores, EMS sales managers noticed that inner soles were moving very briskly in specialty stores. These stores had perfected a multi-step sales technique that included the recommendation of socks designed for specific uses, such as running or hiking, along with an inner sole custom-fit to each customer. Wikis and blogs made it easier for managers to discuss this tactic and share it with the rest of the retail network.

Longer term, EMS is planning for more detailed interactions with its suppliers. By sharing inventory and sales data with suppliers, EMS will be able to quickly restock inventory to meet customer demand, while suppliers will know when to ramp up production.

Eastern Mountain Sports’ executive dashboards are a powerful illustration of how information systems improve decision making. Management was unable to make good decisions about how and where to stock stores because the required data were scattered in many different systems and were difficult to access. Management reporting was excessively manual. Bad decisions about how to stock stores and warehouses increased operating costs and prevented EMS stores from responding quickly to customer needs.

EMS management could have continued to use its outdated management reporting system or implemented a large-scale enterprise-wide database and software, which would have been extremely expensive and time-consuming to complete. Instead, it opted for a business intelligence solution that could extract, consolidate, and analyze sales and merchandising data from its various legacy systems. It chose a platform from Information Builders because the tools were user-friendly and capable of pulling together data from many different sources.

The chosen solution populates a data mart with data from point-of-sale and legacy systems and then pulls information from the data mart into a central series of executive dashboards visible to authorized users throughout the organization. Decision-makers are able to quickly access a unified high-level view of key performance indicators such as sales, inventory, and margin levels or drill down to obtain more detail about specific transactions. Increased availability of this information has helped EMS managers make better decisions about increasing sales, allocating resources, and propagating best practices.

12.1 Decision Making and Information Systems

Decision making in businesses used to be limited to management. Today, lower-level employees are responsible for some of these decisions, as information systems make information available to lower levels of the business. But what do we mean by better decision making? How does decision making take place in businesses and other organizations? Let’s take a closer look.

BUSINESS VALUE OF IMPROVED DECISION MAKING

What does it mean to the business to make better decisions? What is the monetary value of improved decision making? Table 12-1 attempts to measure the monetary value of improved decision making for a small U.S. manufacturing firm with $280 million in annual revenue and 140 employees. The firm has identified a number of key decisions where new system investments might improve the quality of decision making. The table provides selected estimates of annual value (in the form of cost savings or increased revenue) from improved decision making in selected areas of the business.

We can see from Table 12-1 that decisions are made at all levels of the firm and that some of these decisions are common, routine, and numerous. Although the value of improving any single decision may be small, improving hundreds of thousands of “small” decisions adds up to a large annual value for the business.

TYPES OF DECISIONS

Chapters 1 and 2 showed that there are different levels in an organization. Each of these levels has different information requirements for decision support and responsibility for different types of decisions (see Figure 12-1). Decisions are classified as structured, semistructured, and unstructured.

TABLE 12-1 BUSINESS VALUE OF ENHANCED DECISION MAKING

EXAMPLE DECISION

DECISION MAKER

NUMBER OF ANNUAL DECISIONS

ESTIMATED VALUE TO FIRM OF A SINGLE IMPROVED DECISION

ANNUAL VAUE

Allocate support to most valuable customers

Accounts manager

12

$ 100,000

$1,200,000

Predict call center daily demand

Call center management

4

150,000

600,000

Decide parts inventory levels daily

Inventory manager

365

5,000

1,825,000

Identify competitive bids from major suppliers

Senior management

1

2,000,000

2,000,000

Schedule production to fill orders

Manufacturing manager

150

10,000

1,500,000

Allocate labor to complete a job

Production floor manager

100

4,000

400,000

FIGURE 12-1 INFORMATION REQUIREMENTS OF KEY DECISION-MAKING GROUPS IN A FIRM

Senior managers, middle managers, operational managers, and employees have different types of decisions and information requirements.

Unstructured decisions are those in which the decision maker must provide judgment, evaluation, and insight to solve the problem. Each of these decisions is novel, important, and nonroutine, and there is no well-understood or agreed-on procedure for making them.

Structured decisions, by contrast, are repetitive and routine, and they involve a definite procedure for handling them so that they do not have to be treated each time as if they were new. Many decisions have elements of both types of decisions and are semistructured, where only part of the problem has a clear-cut answer provided by an accepted procedure. In general, structured decisions are more prevalent at lower organizational levels, whereas unstructured problems are more common at higher levels of the firm.

Senior executives face many unstructured decision situations, such as establishing the firm’s five- or ten-year goals or deciding new markets to enter. Answering the question “Should we enter a new market?” would require access to news, government reports, and industry views as well as high-level summaries of firm performance. However, the answer would also require senior managers to use their own best judgment and poll other managers for their opinions.

Middle management faces more structured decision scenarios but their decisions may include unstructured components. A typical middle-level management decision might be “Why is the reported order fulfillment report showing a decline over the past six months at a distribution center in Minneapolis?” This middle manager will obtain a report from the firm’s enterprise system or distribution management system on order activity and operational efficiency at the Minneapolis distribution center. This is the structured part of the decision. But before arriving at an answer, this middle manager will have to interview employees and gather more unstructured information from external sources about local economic conditions or sales trends.

Operational management and rank-and-file employees tend to make more structured decisions. For example, a supervisor on an assembly line has to decide whether an hourly paid worker is entitled to overtime pay. If the employee worked more than eight hours on a particular day, the supervisor would routinely grant overtime pay for any time beyond eight hours that was clocked on that day.

A sales account representative often has to make decisions about extending credit to customers by consulting the firm’s customer database that contains credit information. If the customer met the firm’s prespecified criteria for granting credit, the account representative would grant that customer credit to make a purchase. In both instances, the decisions are highly structured and are routinely made thousands of times each day in most large firms. The answer has been preprogrammed into the firm’s payroll and accounts receivable systems.

THE DECISION-MAKING PROCESS

Making a decision is a multistep process. Simon (1960) described four different stages in decision making: intelligence, design, choice, and implementation (see Figure 12-2).

FIGURE 12-2 STAGES IN DECISION MAKING

The decision-making process is broken down into four stages.

Intelligence consists of discovering, identifying, and understanding the problems occurring in the organization—why a problem exists, where, and what effects it is having on the firm.

Design involves identifying and exploring various solutions to the problem.

Choice consists of choosing among solution alternatives.

Implementation involves making the chosen alternative work and continuing to monitor how well the solution is working.

What happens if the solution you have chosen doesn’t work? Figure 12-2 shows that you can return to an earlier stage in the decision-making process and repeat it if necessary. For instance, in the face of declining sales, a sales management team may decide to pay the sales force a higher commission for making more sales to spur on the sales effort. If this does not produce sales increases, managers would need to investigate whether the problem stems from poor product design, inadequate customer support, or a host of other causes that call for a different solution.

MANAGERS AND DECISION MAKING IN THE REAL WORLD

The premise of this book and this chapter is that systems to support decision making produce better decision making by managers and employees, above average returns on investment for the firm, and ultimately higher profitability. However, information systems cannot improve all the different kinds of decisions taking place in an organization. Let’s examine the role of managers and decision making in organizations to see why this is so.

Managerial Roles

Managers play key roles in organizations. Their responsibilities range from making decisions, to writing reports, to attending meetings, to arranging birthday parties. We are able to better understand managerial functions and roles by examining classical and contemporary models of managerial behavior.

The classical model of management, which describes what managers do, was largely unquestioned for the more than 70 years since the 1920s. Henri Fayol and other early writers first described the five classical functions of managers as planning, organizing, coordinating, deciding, and controlling. This description of management activities dominated management thought for a long time, and it is still popular today.

The classical model describes formal managerial functions but does not address what exactly managers do when they plan, decide things, and control the work of others. For this, we must turn to the work of contemporary behavioral scientists who have studied managers in daily action. Behavioral models state that the actual behavior of managers appears to be less systematic, more informal, less reflective, more reactive, and less well organized than the classical model would have us believe.

Observers find that managerial behavior actually has five attributes that differ greatly from the classical description. First, managers perform a great deal of work at an unrelenting pace—studies have found that managers engage in more than 600 different activities each day, with no break in their pace. Second, managerial activities are fragmented; most activities last for less than nine minutes, and only 10 percent of the activities exceed one hour in duration. Third, managers prefer current, specific, and ad hoc information (printed information often will be too old). Fourth, they prefer oral forms of communication to written forms because oral media provide greater flexibility, require less effort, and bring a faster response. Fifth, managers give high priority to maintaining a diverse and complex web of contacts that acts as an informal information system and helps them execute their personal agendas and short- and long-term goals.

Analyzing managers’ day-to-day behavior, Mintzberg found that it could be classified into 10 managerial roles. Managerial roles are expectations of the activities that managers should perform in an organization. Mintzberg found that these managerial roles fell into three categories: interpersonal, informational, and decisional.

Interpersonal Roles. Managers act as figureheads for the organization when they represent their companies to the outside world and perform symbolic duties, such as giving out employee awards, in their interpersonal role. Managers act as leaders, attempting to motivate, counsel, and support subordinates. Managers also act as liaisons between various organizational levels; within each of these levels, they serve as liaisons among the members of the management team. Managers provide time and favors, which they expect to be returned.

Informational Roles. In their informational role, managers act as the nerve centers of their organizations, receiving the most concrete, up-to-date information and redistributing it to those who need to be aware of it. Managers are therefore information disseminators and spokespersons for their organizations.

Decisional Roles. Managers make decisions. In their decisional role, they act as entrepreneurs by initiating new kinds of activities; they handle disturbances arising in the organization; they allocate resources to staff members who need them; and they negotiate conflicts and mediate between conflicting groups.

Table 12-2, based on Mintzberg’s role classifications, is one look at where systems can and cannot help managers. The table shows that information systems do not yet contribute to some important areas of management life.

TABLE 12-2 MANAGERIAL ROLES AND SUPPORTING INFORMATION SYSTEMS

Sources: Kenneth C. Laudon and Jane P. Laudon; and Mintzberg, 1971.

Real-World Decision Making

We now see that information systems are not helpful for all managerial roles. And in those managerial roles where information systems might improve decisions, investments in information technology do not always produce positive results. There are three main reasons: information quality, management filters, and organizational culture (see Chapter 3).

Information Quality. High-quality decisions require high-quality information. Table 12-3 describes information quality dimensions that affect the quality of decisions.

If the output of information systems does not meet these quality criteria, decision-making will suffer. Chapter 6 has shown that corporate databases and files have varying levels of inaccuracy and incompleteness, which in turn will degrade the quality of decision making.

Management Filters. Even with timely, accurate information, some managers make bad decisions. Managers (like all human beings) absorb information through a series of filters to make sense of the world around them. Managers have selective attention, focus on certain kinds of problems and solutions, and have a variety of biases that reject information that does not conform to their prior conceptions.

For instance, Wall Street firms such as Bear Stearns and Lehman Brothers imploded in 2008 because they underestimated the risk of their investments in complex mortgage securities, many of which were based on subprime loans that were more likely to default. The computer models they and other financial institutions used to manage risk were based on overly optimistic assumptions and overly simplistic data about what might go wrong. Management wanted to make sure that their firms’ capital was not all tied up as a cushion against defaults from risky investments, preventing them from investing it to generate profits. So the designers of these risk management systems were encouraged to measure risks in a way that did not pick them all up. Some trading desks also oversimplified the information maintained about the mortgage securities to make them appear as simple bonds with higher ratings than were warranted by their underlying components (Hansell, 2008).

Organizational Inertia and Politics. Organizations are bureaucracies with limited capabilities and competencies for acting decisively. When environments change and businesses need to adopt new business models to survive, strong forces within organizations resist making decisions calling for major change. Decisions taken by a firm often represent a balancing of the firm’s various interest groups rather than the best solution to the problem.

TABLE 12-3 INFORMATION QUALITY DIMENSIONS

QUALITY DIMENSION

DESCRIPTION

Accuracy

Do the data represent reality?

Integrity

Are the structure of data and relationships among the entities and attributes consistent?

Consistency

Are data elements consistently defined?

Completeness

Are all the necessary data present?

Validity

Do data values fall within defined ranges?

Timeliness

Area data available when needed?

Accessibility

Are the data accessible, comprehensible, and usable?

Studies of business restructuring find that firms tend to ignore poor performance until threatened by outside takeovers, and they systematically blame poor performance on external forces beyond their control such as economic conditions (the economy), foreign competition, and rising prices, rather than blaming senior or middle management for poor business judgment (John, Lang, Netter, et al., 1992).

12.2 Systems for Decision Support

There are four kinds of systems for supporting the different levels and types of decisions we have just described. We introduced some of these systems in Chapter 2. Management information systems (MIS) provide routine reports and summaries of transaction-level data to middle and operational level managers to provide answers to structured and semistructured decision problems. Decision-support systems (DSS) provide analytical models or tools for analyzing large quantities of data for middle managers who face semistructured decision situations. Executive support systems (ESS) are systems that provide senior management, making primarily unstructured decisions, with external information (news, stock analyses, and industry trends) and high-level summaries of firm performance.

In this chapter, you’ll also learn about systems for supporting decision-makers working as a group. Group decision-support systems (GDSS) are specialized systems that provide a group electronic environment in which managers and teams are able to collectively make decisions and design solutions for unstructured and semistructured problems.

MANAGEMENT INFORMATION SYSTEMS (MIS)

MIS, which we introduced in Chapter 2 help managers monitor and control the business by providing information on the firm’s performance. They typically produce fixed, regularly scheduled reports based on data extracted and summarized from the firm’s underlying transaction processing systems (TPS). Sometimes, MIS reports are exception reports, highlighting only exceptional conditions, such as when the sales quotas for a specific territory fall below an anticipated level or employees have exceeded their spending limits in a dental care plan. Today, many of these reports are available online through an intranet, and more MIS reports are generated on demand. Table 12-4 provides some examples of MIS applications.

DECISION-SUPPORT SYSTEMS (DSS)

Whereas MIS primarily address structured problems, DSS support semistructured and unstructured problem analysis. The earliest DSS were heavily model-driven, using some type of model to perform “what-if” and other kinds of analyses. Their analysis capabilities were based on a strong theory or model combined with a good user interface that made the system easy to use. The voyage-estimating DSS and Air Canada maintenance system described in Chapter 2 are examples of model-driven DSS.

TABLE 12-4 EXAMPLES OF MIS APPLICATIONS

COMPANY

MIS APPLICATION

California Pizza Kitchen

Inventory Express application “remembers” each restaurant’s ordering patterns and compares the amount of ingredients used per menu item to predefined portion measurements established by management. The system identifies restaurants with out-of-line portions and notifies their managers so that corrective actions will be taken.

PharMark

Extranet MIS identifies patients with drug-use patterns that place them at risk for adverse outcomes.

Black & Veatch

Intranet MIS tracks construction costs for various projects across the United States.

Taco Bell

Total Automation of Company Operations (TACO) system provides information on food, labor, and period-to-date costs for each restaurant.

The Interactive Session on Management describes another model-driven DSS. In this particular case, the system did not perform as well as expected because of the assumptions driving the model and user efforts to circumvent the system. As you read this case, try to identify the problem this company was facing, what alternative solutions were available to management, and how well the chosen solution worked.

Some contemporary DSS are data-driven, using online analytical processing (OLAP), and data mining to analyze large pools of data. The business intelligence applications described in Chapter 6 are examples of these data-driven DSS, as are the spreadsheet pivot table applications we describe in this section. Data-driven DSS support decision making by enabling users to extract useful information that was previously buried in large quantities of data. The Interactive Session on Technology provides an example.

Components of DSS

Figure 12-3 illustrates the components of a DSS. They include a database of data used for query and analysis; a software system with models, data mining, and other analytical tools; and a user interface.

The DSS database is a collection of current or historical data from a number of applications or groups. It may be a small database residing on a PC that contains a subset of corporate data that has been downloaded and possibly combined with external data. Alternatively, the DSS database may be a massive data warehouse that is continuously updated by major corporate TPS (including enterprise applications) and data generated by Web site transactions). The data in DSS databases are generally extracts or copies of production databases so that using the DSS does not interfere with critical operational systems.

The DSS user interface permits easy interaction between users of the system and the DSS software tools. Many DSS today have Web interfaces to take advantage of graphical displays, interactivity, and ease of use.

The DSS software system contains the software tools that are used for data analysis. It may contain various OLAP tools, data mining tools, or a collection of mathematical and analytical models that are accessible to the DSS user. A model is an abstract representation that illustrates the components or relationships of a phenomenon. A model may be a physical model (such as a model airplane), a mathematical model (such as an equation), or a verbal model (such as a description of a procedure for writing an order).

INTERACTIVE SESSION: MANAGEMENT TOO MANY BUMPED FLIERS: WHY?

In a seemingly simpler and less hectic time, overbooked flights presented an opportunity. Frequent travelers regularly and eagerly chose to give up their seats and delay their departures by a few hours in exchange for rewards such as a voucher for a free ticket.

Today, fewer people are volunteering to give up their seats for a flight because there are fewer and fewer seats to be bumped to. Airlines are struggling to stay in business and look to save costs wherever possible. They are scheduling fewer flights and those flights are more crowded. Instead of delaying his or her trip by a few hours, a passenger that accepts a voucher for being bumped may have to wait several days before a seat becomes available on another flight. And passengers are being bumped from flights involuntarily more often.

Airlines routinely overbook flights to compensate for the millions of no-shows that cut into expected revenue. The purpose of overbooking is not to leave passengers without a seat, but to come as close as possible to filling every seat on every flight. The revenue lost from an empty seat is much greater than the costs of compensating a bumped passenger. Airlines are much closer today to filling every seat on flights than at any point in their history. The problem is, the most popular routes often sell out, so bumped passengers may be stranded for days.

The airlines do not approach overbooking haphazardly. They employ young, sharp minds with backgrounds in math and economics as analysts. The analysts use computer modeling to predict how many passengers will fail to show up for a flight. They recommend overbooking based on the numbers generated by the software.

The software used by US Airways, for example, analyzes the historical record of no-shows on flights and looks at the rate at every fare level available. The lowest-priced fares are generally nonrefundable, and passengers at those fare levels tend to carry their reservations through. Business travelers with the high-priced fares no-show more often. The software examines the fares people are booking on each upcoming flight and takes other data into account, such as the rate of no-shows on flights originating from certain geographic regions. Analysts then predict the number of no-shows on a particular flight, based on which fares passengers have booked, and overbook the flight accordingly.

Of course, the analysts don’t always guess correctly. And their efforts may be hampered by a number of factors. Ticket agents report that faulty computer algorithms result in miscalculations. Changes in weather can introduce unanticipated weight restrictions. Sometimes a smaller plane is substituted for the scheduled plane. All of these circumstances result in fewer seats being available for the same number of passengers, which might have been set too high already.

Regardless of how much support the analysts have from airline management, gate attendants complain because they are the ones who receive the brunt of overbooked passengers’ wrath. Attendants have been known to call in sick to avoid dealing with the havoc caused by overbooked flights.

Some gate attendants have gone as far as creating phony reservations, sometimes in the names of airline executives or cartoon characters, such as Mickey Mouse, in an effort to stop analysts from overbooking. This tactic may save the attendants some grief in the short term, but their actions often come back to haunt them. The modeling software counts the phony reservations as no-shows, which leads the analysts to overbook the flight even more the next time. Thomas Trenga, vice president for revenue management at US Airways, refers to this game of chicken as “the death spiral.” US Airways discourages the practice of entering phony reservations.

With fewer passengers volunteering to accept vouchers, tensions often escalate. The number of passengers bumped involuntarily in 2006 rose 23 percent from the previous year and has continued to rise. The encouraging statistic is that only 676,408 of the 555 million people who flew in 2006 were bumped, voluntarily or involuntarily.

W. Douglas Parker, CEO of US Airways, said that airlines have to overbook their flights as long as they allow passengers to no-show without penalty. US Airways has a no-show rate of between 7 and 8 percent. US Airways claimed that overbooking contributed to at least $1 billion of its 2006 revenue of $11.56 billion. With a profit of only $304 million, that extra revenue was critical to the survival of the business. Some airlines, such as JetBlue, have avoided the overbooking controversy by offering only nonrefundable tickets. No-shows cannot reclaim the price of their tickets. Business travelers often buy the most expensive seats, but also want the flexibility of refundable tickets, so JetBlue is considering a change in its policy.

The airlines are supposed to hold their analysts accountable for their work, but they are rarely subject to critical review. Some analysts make an effort to accommodate the wishes of the airport workers by finding a compromise in the overbooking rate. Unfortunately, analysts often leave their jobs for new challenges once they become proficient at overbooking.

Sources: Dean Foust and Justin Bachman, “You Think Flying Is Bad Now...,” Business Week, May 28, 2008; “The Unfriendly Skies,” USA Today, June 4, 2008; Jeff Bailey, “Bumped Fliers and Plan B,” The New York Times, May 30, 2007; and Alice LaPlante, “Travel Problems? Blame Technology,” InformationWeek.com, June 11, 2007.

CASE STUDY QUESTIONS

1. Is the decision support system being used by airlines to overbook flights working well? Answer from the perspective of the airlines and from the perspective of customers.

2. What is the impact on the airlines if they are bumping too many passengers?

3. What are the inputs, processes, and outputs of this DSS?

4. What people, organization, and technology factors are responsible for excessive bumping problems?

5. How much of this is a “people” problem? Explain your answer.

MIS IN ACTION

Visit the Web sites for US Airways, JetBlue, and Continental. Search the sites to answer the following questions:

1. What is the policy of each of these airlines for dealing with involuntary refunds (overbookings)? (Hint: These matters are often covered in the Contract of Carriage.)

2. In your opinion, which airline has the best policy? What makes this policy better than the others?

3. How are each of these policies intended to benefit customers? How do they benefit the airlines?

FIGURE 12-3 OVERVIEW OF A DECISION-SUPPORT SYSTEM

The main components of the DSS are the DSS database, the user interface, and the DSS software system. The DSS database may be a small database residing on a PC or a large data warehouse.

INTERACTIVE SESSION: TECHNOLOGY BUSINESS INTELLIGENCE TURNS DICK’S SPORTING GOODS INTO A WINNER

Dick’s Sporting Goods is a prominent retailer of sporting apparel and equipment based primarily in the eastern half of the United States. The company was founded in 1948 by Dick Stack, who was only 18 years old at the time. Stack’s business initially sold only fishing supplies, but gradually expanded to sell general sporting goods. In the 1990s, under the stew-ardship of Stack’s son Ed, the retailer began rapid growth in an effort to become a national sporting goods chain. Today, Dick’s operates over 300 stores in 34 states and earns annual revenue of just under $4 billion. It also owns Golf Galaxy, a golf specialty retailer. The company planned to add 44 new stores in 2008 and has maintained a strong position during difficult economic conditions.

Dick’s has flourished because it focuses on being an authentic sporting goods retailer by offering a broad selection of high-quality, competitively priced brand-name sporting goods equipment, apparel, and footwear that enhances its customers’ performance and enjoyment of their sports activities. However, Dick’s has had problems managing its inventory and making decisions about how to stock its stores. These problems stemmed from outdated merchandise management software and threatened to curtail Dick’s lofty plans for the future.

The company initially used a merchandising system from STS as a basic reporting tool. The system wasn’t well suited to the needs of the company. It was able to compile sales figures for athletic gear and clothing, but it wasn’t able to analyze how a specific item, such as a Wilson Tennis n4 racquet, was performing regionally or in a particular store. Instead, it automatically aggregated information from all stores and combined it into a single report. Retrieving data from the database was a long, inefficient process, sometimes taking over an hour to complete, and wasn’t satisfactory for answering questions requiring complex analysis.

Because there was no central repository for company information, it was also difficult to tell whether or not a particular report was accurate. There were no standard company-wide sales and inventory reports. The company lacked a unified database that all of the company’s employees could access. Employees kept their own analyses of sales and inventory in their own departments and on their own machines. Sometimes they lost their reports because they did not remember the names of their data files. Recognizing the problems, Dick’s attempted to roll out new tools intended to update the company’s data storage and information retrieval processes. But employees resisted the change, preferring the methods they were used to over new tools from Cognos, a maker of business intelligence software.

Dick’s decided to perform a complete overhaul of their data storage system in 2003. The new system featured software from MicroStrategy and a database from Oracle. The database Dick’s selected was Oracle’s 8i database with customized capabilities to extract data and the ability to transform to meet different business requirements. It has since been upgraded to a more advanced 10g model. The new system was able to track the sale of apparel and equipment in each store and by region.

The new system was launched with a training program to promote user adoption, so that employees didn’t persist in using the old system that they were more accustomed to. Even with the new training system, employee adoption was slow, but the company offered incentives to using the new system and gradually phased out the old one. Only when the old system was phased out completely did adoption of the new system increase tenfold. Some of the failings of previous information systems were attributed to lack of training programs to smooth the difficulties of adopting new systems, and this time around Dick’s ensured that the proper programs were in place.

The MicroStrategy software was a key element of Dick’s overhaul. What sets MicroStrategy apart from competing products is its ability to work with relational databases via relational online analytical processing (ROLAP). Multidimensional OLAP uses a multidimensional database for analysis (see Chapter 6), whereas ROLAP accesses data directly from data warehouses. It dynamically consolidates data for ad hoc and decision support analyses and scales to a large number of business analysis perspectives (dimensions) while MOLAP generally performs efficiently with 10 or fewer dimensions. The software allows Dick’s employees to perform detailed analyses to track sales and inventory levels.

MicroStrategy allows Dick’s employees to create different types of reports. For example, ‘canned’ reports are reports with settings frequently used by other employees. If an employee needs a report with commonly requested parameters, canned reports save workers the time and energy required to expressly set those parameters. On the other hand, ‘self-service’ reports have customized inputs and outputs for instances when a unique piece of information is sought. Processes that once took hours now take mere minutes because of the system’s interaction with the master database, which consists of multiple terabytes of data.

Recent results suggest that the implementation has paid off for Dick’s, as their earnings have doubled since the initiative began and their operating margin has been close to double that of their competitors going forward. Sales in Q1 2008 were up 11 percent to $912 million, and although the company hasn’t been immune to the difficult economic conditions, the company is outperforming its competitors and has its sights set on gaining market share during the downturn. Although the company’s stock price has not reached levels that it was expected to in the past several years, the company’s future outlook remains positive, in large part due to their successful IT implementation.

Sources: MicroStrategy, “Success Story: Dick’s Sporting Goods Inc.,” 2008; Brian P. Watson, “Business Intelligence: Will It Improve Inventory?” www.baselinemag.com, May 14, 2007; “Dick’s Sporting Goods Form 10-K Annual Report,” March 27, 2008; “Dick’s Sporting Goods Inc., Q1 2008 Earnings Call Transcript,” www.seekingalpha.com, May 22, 2008.

CASE STUDY QUESTIONS

1. What problems did Dick’s face with its data tracking and reporting? How did they affect decision making and business performance?

2. What did the company do to remedy those problems?

3. Was MicroStrategy an appropriate selection for Dick’s? Why or why not?

4. Has improved reporting solved all of this company’s problems? Explain your answer.

MIS IN ACTION

Explore the MicroStrategy Web site and then answer the following questions:

1. Describe the capabilities of MicroStrategy software. List the capabilities that would be most useful for supporting decisions about stocking Dick’s stores. Explain how the software would help Dick’s employees with these decisions.

2. Review the section on MicroStrategy’s Dynamic Enterprise Dashboards. Then design a dashboard for a manager deciding how to stock Dick’s stores.

Statistical modeling helps establish relationships, such as relating product sales to differences in age, income, or other factors between communities. Optimization models determine optimal resource allocation to maximize or minimize specified variables, such as cost or time. A classic use of optimization models is to determine the proper mix of products within a given market to maximize profits.

Forecasting models often are used to forecast sales. The user of this type of model might supply a range of historical data to project future conditions and the sales that might result from those conditions. The decision maker could vary those future conditions (entering, for example, a rise in raw materials costs or the entry of a new, low-priced competitor in the market) to determine how new conditions might affect sales.