Turning Knowledge Into Power: Using Business Intelligence to Improve Decision-Making

Better Business Intelligence.


Most businesses are sitting on a gold mine of data but lack tools to mine the wealth. Why aren’t companies better-using business intelligence, which receives so much media attention, to improve decision-making and work processes? The overriding problem is the quality of data and its accessibility. Here’s how savvy companies are using the tools of business intelligence to empower the enterprise.


Business Intelligence Tools


Online Analytical Processing (OLAP): Tools that create a structure known as a data cube allows users to quickly drill down into data to analyze trends and view patterns. A type of decision-support system, OLAP lets users quickly gain access to summarized, multidimensional data through a Web browser. It is particularly adept at handling financial summaries.


Relational OLAP (ROLAP): A tool that extracts data from conventional relational databases, using SQL statements. ROLAP can be used through a Web browser.


Multidimensional OLAP (MOLAP): Also referred to as OLAP. Summarizes transactions into multidimensional views.


Business Intelligence Portal: A corporate portal, accessible via a Web browser, that allows users to query and produce reports using enterprisewide data.


Data Mining: Exploring detailed business transactions manually or through automated systems connected to a data warehouse or other source of data.


Decision-Support System (DSS): An information and planning system that allows ad-hoc queries and can analyze and predict the impact of potential decisions.


Executive Information System (EIS): An information system that consolidates and summarizes transactions within the organization. An EIS provides management with needed information from internal and external sources. An EIS that provides “what if?” capabilities is also considered a decision-support system.


If you ask Dave Anderson what his biggest business challenge is, the CIO of the systems division for ITT Industries in Colorado Springs, Colo., one of the nation’s leading defense contractors, will tell you that it’s keeping costs down and operating as efficiently as possible. And if you ask him how the $450 million company does it, he will answer with a simple explanation: business intelligence — the use of software and systems to improve the decision-making process within an organization. “If your goal is to reduce labor and purchasing costs, and understand them in context with a particular project, you must sort through a mountain of data. Without the right tools there’s no way to drill down and find the essential information.”


For ITT, which manages a slew of high-security government facilities including the Pacific Missile Range Facility in Hawaii and the Cheyenne Mountain Air Station in Colorado, business intelligence has become an indispensable tool in the corporate arsenal. It lets managers, including those in the finance department, drill down through layers of data to discover patterns that otherwise wouldn’t be visible. Using software and a standard Web browser, “we are able to sift through all the data and spot patterns that seem unusual but can’t always be defined up front,” says Anderson.


Spotting patterns includes activities such as identifying potential cost overruns and recognizing a product development cycle that is out of sync. It can also mean slicing through financial data to spot patterns of inefficient labor use. Because the business intelligence data is contained in a single database in Colorado Springs but is available at more than 20 company locations, information that used to be squirreled away in silos and unavailable at the enterprise level is now at the center of an information revolution. “We are able to run the business at a far more strategic level,” Anderson says.


Business intelligence is certainly not a new concept. For decades, savvy executives have found ways to outsmart the competition — through lower prices, better service, and more efficient manufacturing or business processes. In many cases, they have pored over ledgers and spreadsheets in order to spot trends that could be transformed into specific actions. And as early as the 1970s when computers entered the picture, management information systems (MIS) departments attempted to aggregate and organize data — often unsuccessfully.


However, in today’s electronic workplace, where systems and software manage vast reservoirs of data, business intelligence (BI) is changing the enterprise at leading-edge organizations. Business intelligence encompasses a broad category of applications and technologies for gathering, storing, analyzing and accessing data, and can tackle everything from enterprisewide data analysis to the highly specialized requirements of a department or workgroup. Decision support, query and reporting, online analytical processing (OLAP), statistical analysis, forecasting, and data mining are just some of the tools that let managers make highly focused business decisions. In other words, business intelligence involves virtually any technology or tool that aggregates, manipulates and formats data into reports.


The Competitive Edge

Business intelligence can makes things happen.”The ability to truly understand markets, competitors and processes are extremely powerful.” says Glen Marianko, president and chief technologist at Progressive Strategies Inc., a New York-based consulting firm. According to IBM Corp., a typical company harnesses only 2 percent to 4 percent of the data that resides in its systems. The rest sits in databases and is never touched. The Futures Group, a Boston-based consulting firm, has found that only about 60 percent of large companies use business intelligence. Richard W. Lewis, senior vice president at QLV Goal Ltd., a Schomberg, Ontario, consultancy specializing in competitive information management, states, “Most companies are sitting on a gold mine but have no way to mine the gold.”


Data mining and analysis tools now make it possible to cull, aggregate and drill through data in ways that previously didn’t exist. But business intelligence doesn’t happen in a vacuum. It requires back-end tools — databases, data warehouses, datamarts and middleware — that can store and organize information in a way that makes in-depth analysis and queries possible. On the front end, it means putting powerful query tools and analytic capabilities in the hands of workers, so they can dissect data and make highly informed decisions. “It’s all about fully exploiting data, information and knowledge,” says Dwight Davis, service director at Summit Strategies in Kirkland, Wash.


Richard Philyaw, an assistant vice president of architecture and strategy at interBiz Financial Group, a division of software firm Computer Associates in Islandia, N.Y., says that business intelligence addresses three broad functional areas. First, it facilitates collaboration between businesses via the Web. Second, it provides the tools to “intelligently analyze and visualize the extended enterprise in real time.” Third, it helps predict future business events and trends. “Business intelligence can seamlessly bring together disparate information sources,” he says. What’s more, it can fit into other organizational initiatives, such as enterprise resource planning (ERP) systems or portals — electronic gateways to information inside and outside of companies.


In fact, when all of the pieces come together, it’s suddenly possible to gain insight into manufacturing processes, inventory controls, sales, marketing, financials, customer relationships, human resources and more. It’s possible to develop efficient just-in-time (JIT) methods, use supply-chain-management techniques, and anticipate and avoid problems before they occur. “The key,” says Davis, “is accessing, manipulating and organizing data in a way that’s actually useful to your company. While everyone recognizes the importance of business intelligence, many companies find themselves intimidated by the complexity of making it all work.”


ITT Industries has stored a diverse array of data, including human resources (HR) records, operational costs, purchasing costs, budgets and more, and its software lets authorized managers tap into sophisticated neural net capabilities. Using the firm’s intranet, they can drill down through mountains of data stored on a server containing a huge relational database. In minutes, a manager can have a report, chart or graph in hand. “One of the problems in the past is that a group of engineers might conduct detailed research and get off track, but we would have no way to know,” explains Anderson. Now the software can spot anomalies and tip off supervisors. “We can see exactly what’s needed to make the project a success.”

Having used general business intelligence capabilities, the company is now venturing deeper into the technology. HR, for example, is tracking turnover rates, salaries, hits on recruitment pages and job applications to constantly tweak pay and adjust job functions. The system also keeps tabs on sick days, output and other productivity measures so that HR can recognize when project-specific problems exist and take quick action. “Sometimes it shows us a trend that we would never have noticed,” Anderson explains.


Overcoming the Obstacles

Lewis says that when business intelligence is used correctly, it can provide insight that revolutionizes the way work gets done. It’s suddenly possible to predict which customers a company is likely to lose or when the main control board in an ATM machine is likely to fail — and have the right parts on the right truck to fix the problem immediately. It’s also possible to analyze a company’s cash conversion cycle and understand exactly where bottlenecks occur and how they throw off the entire manufacturing process.


Yet putting all the pieces together isn’t simple. The overriding problem for most companies attempting to embrace business intelligence, Lewis insists, is the quality of the data and its accessibility, which can make or break a BI initiative. “Data acquisition, interaction with legacy systems, and data cleansing and manipulation are all fundamental problems for many organizations,” he explains.


Adding to the challenge, says Davis, is growing access to inexpensive, nearly unlimited storage. Today, it’s easy to archive terabytes of data, though it’s far more vexing to connect systems and get to the data that’s required to make a key business decision. “Many companies don’t know what information they have, and they certainly don’t know what’s valuable and what isn’t,” he points out. What’s more, “most data isn’t stored neatly in a single SQL database; it’s located across file systems, e-mail systems and different kinds of databases.” Extracting, reformatting or converting it — and getting it into a data warehouse or connected by middleware — can be extremely difficult.


Yet when a business intelligence system is designed correctly, Lewis says, it can lead to real-time collaborative decision support. Quite simply, “you’re taking information, knowledge and expertise that already exists and creating a just-in-time model for data delivery. At that point, it’s possible to build highly efficient business delivery systems, business marketing systems and true value chains,” he explains. What’s more, Web-based business intelligence isn’t limited by traditional boundaries or barriers. The Internet can create entirely new processes and solutions.


At Republic Indemnity Co. of America, an Encino, Calif., provider of workers’ compensation insurance, managers can sift through a combination of internal and external data in order to better understand pricing for different states. Using highly specialized data-mining tools, employees can understand which markets are profitable, which aren’t and how to adjust pricing for optimal profits. Those using the business intelligence tools can tap into more than 30 years of the firm’s performance data, along with detailed demographic and financial data from various states.


After deregulation hit the California insurance market in 1996, Republic Indemnity first turned to Microsoft Access and Excel to examine the data. Unfortunately, with millions of records packed into its systems, managers quickly felt overwhelmed by the process. Now “we’re able to determine which products are profitable in which markets, and instantly adjust pricing to maximize the profit potential for that market or simply become competitive in it,” says Bob Cancilla, director of systems planning and administration. “In many industries, companies can no longer expect to collect money simply because they’re in the market. Survival depends on using business intelligence tools to compete more effectively.”


People and Processes

Summit Strategies’ Davis believes that many vendor solutions — including those that provide sophisticated online analytical processing (OLAP) and relational OLAP (ROLAP) tools to manage public applications and shared views of libraries — present remarkable opportunities. Nevertheless, business intelligence is just as much about people and processes as technology. “Without the right human systems in place, business intelligence can be an expensive proposition with a questionable return on investment.”


With a 9-terabyte datamart, banking giant Wells Fargo could find itself buried in data. In fact, the bank has recorded every customer transaction for the last two-and-a-half years and slotted them into an Oracle Corp. database, complete with time and date stamps. David Holvey, senior VP of the consumer banking group, Wells Fargo in San Francisco, uses clustering algorithms, trees, nests and other methods to conduct complex predictions of behavior. He can examine what products customers are likely to use, when they are likely to need them, and when they’re likely to look for another bank altogether.


“Today’s business intelligence tools have evolved from complex systems that put a high demand on users’ knowledge to software that doesn’t require advanced programming or skills,” says Holvey. But that’s a mixed blessing, he believes. Although ease of use and greater accessibility puts business intelligence into the hands of a growing number of companies — many of which are using the tools to strategic advantage — it can also create more than its fair share of problems. “One of the dangers is that you can use these tools to conduct analysis that doesn’t make sense. You can wind up generating results that aren’t valid and coming to conclusions that aren’t useful.”


Achieving success in business intelligence isn’t easy. A multitude of solutions exists, and a project can take months to implement and cost millions of dollars — with no guarantee of results. Lewis suggests that organizations weigh the alternatives and address critical business needs first. Regardless of the exact approach or the specific tools, it’s possible to open small data streams for specific uses or departments and eventually let them flow together into a river of information.


According to Fred Wergeles, a senior consultant with The Futures Group, Glastonbury, Conn., the goal of business intelligence isn’t collecting everything from everyone. “Well-coordinated collection, analysis and dissemination techniques across the organization are far more likely to result in success,” he says. Moreover, it’s difficult to measure the success of business intelligence only in dollars and cents. Although it can cut costs and boost productivity, say analysts, BI often becomes an inextricable part of an organization. It ripples through the workforce and creates ancillary benefits for the company and the entire value chain.


Business intelligence is finally making it possible to transform huge repositories of data into a strategic resource. When it’s used effectively, “people can discover all sorts of relationships and trends that they generally wouldn’t see,” points out Davis.

Though deploying BI tools requires a tremendous amount of effort, the payoff can mean the difference between prospering and not being in business at all.