If ever there was a company that appears to have its act together when it comes to giving decision-makers the information they need, it's Eastman Chemical Co. The Kingsport, Tenn., chemical maker has an Oracle data warehouse full of information pulled from its E-commerce system and global SAP R/3 applications. Decision-support and business-scorecard apps based on Cognos Inc. software run alongside the data warehouse. The system is supported by a highly standardized infrastructure, including LAN, WAN, and desktop PC software that, in theory, should make data integration and communication a snap.

But Eastman managers don't have easy access to the information they need, says CIO Jerry Hale. That's because there's too much information to cull through, data is in incompatible formats, and the decision-support software is too difficult for business managers to use. And although users can access a number of standard reports, they don't meet the needs of managers who require specific views of the data to do their jobs, Hale says.

Eastman needs to do a better job of delivering those views if the company's IT investments are to pay off, Hale says. "A high percentage of the data is there. The challenge is extracting it and presenting it in a way that's useful for managers to make decisions."

In today's fast-paced environment, businesses need greater access to more information more quickly to make decisions. Are they getting it? Despite collectively spending billions of dollars on data-management and business-intelligence tools, the answer in many cases is no. TruServe Corp. CIO Neil Hastie describes how executives at many companies--including, until recently, his own--make business decisions: "A lot of by-guess and by-golly, a lot of by-gut, and a whole lot of paper reports."

There's more than one explanation for this situation. While companies may collect lots of data, it's often scattered throughout disparate systems and stored in multiple formats that make it difficult to pull together to obtain, say, a single view of a company's customers. Business-intelligence tools for extracting and analyzing information are still too complex for nontechnical people to use, so business managers wind up with reports that are too static or not tailored to their needs. There's often a disconnect between the information business managers need to make day-to-day decisions and what IT provides.

To resolve these problems, companies are turning to a range of technologies, from data warehouses that pull together data from disparate sources to portals and digital dashboards--and even easier-to-use business-intelligence tools--that give managers a snapshot of the information and performance metrics they need to run the business. But while vendors have been providing such software for several years, gaps in packaged apps call for homegrown code, heavy-duty integration, and tight collaboration between business and IT.

The large stores of disparate data often hinder efforts to get the right information to the right decision-makers, says Philip Russom, an independent industry analyst in Waltham, Mass. "The volume of data is increasing, and the number of sources is increasing," he says.

Take the situation at Texas Instruments Inc.'s Kilby semiconductor fabrication plant. The Dallas facility is where TI tests and manufactures new semiconductor products and irons out production and yield problems before manufacturing the chips at high volume. A tremendous amount of performance data is collected from production equipment and prototype semiconductors--data that's used to make chip design and production decisions.

"Data analysis is a very critical part of what we do on a day-to-day basis," says Joe Lebowitz, yield and product engineering director at the plant. But data must be pulled from multiple sources. Semiconductor defect data is in one database, process equipment history (detailing what tools were used to build which materials) is in another, and electrical test data is in yet another. "One of the big tasks at hand is tying those databases together and extracting the data we need," Lebowitz says. Several programmers on his staff do nothing but build data-extraction programs and reports that provide plant managers with needed information. The engineers get the information they need, but at the cost of devoting manpower that could be used to build better chips.

Worthington's data warehouse is designed to give people the best information to let them make decisions, CIO Dove says

Worthington Industries Inc., a Columbus, Ohio, steel-processing company, is designing a data warehouse that combines forecasting, inventory, and sales information from its steel, cylinders, and metal framing operations with data from financial and capacity-planning systems. Planning for the data warehouse began in February, and CIO Jonathan Dove expects to begin deploying the system by year's end. When combined with front-end analysis tools, the system will provide better forecasting capabilities for operational managers and a consolidated view of the company's performance for top executives. "That's our goal: to provide accurate information to the right people to make decisions in a timely manner," Dove says.

But front-end analysis tools can be as much a problem as a solution. While suppliers of business-intelligence applications have been saying for years that their products bring data analysis to front-line knowledge workers, some say a Ph.D. in statistics and experience in SQL programming are all but required to use the software. "These tools were not designed for the average business user," Russom says. That's true for everything from the executive information systems of the early 1990s to most query and reporting tools available today.

In the past, business executives generally delegated the chore of gathering information to their assistants. Now that they have direct access to computerized data, it demands a certain type of software and manager, Russom says. "The self-service thing has been a problem. Many of these people don't have experience in evaluating analytical reports."

Eastman Chemical is exploring the idea of creating a simple menu of reports that managers can choose from; they can then ask to see more details as needed. Eastman may also bring in outside information, such as economic and chemical-industry leading-indicator data, so managers can compare it with internal data to match production and forecasted demand. "Our challenge within the company is to bring this to another level in terms of access and ease of use," Hale says.

Keeping it simple is a challenge at TI's Kilby plant. To access and analyze gigabytes, even terabytes, of data, the plant uses analytical software from SAS Institute Inc., supplemented by a number of custom-built analysis tools and reports. But Lebowitz says the SAS software is too complex. "Analytical software is only as useful as the people using it," he says. "A lot of the power I see in business-analysis tools is getting them into the hands of everyday engineers."

SAS chief technology officer Keith Collins acknowledges that the user interface for SAS's analytical software, originally developed for programmers, isn't the friendliest. In a bid to make the software easier to use, SAS said last week it's developing interfaces tailored to the needs and skill levels of various classes of users, including information consumers, business analysts, executives, and technologists.

Lebowitz's goal is to save TI engineers from having to hunt down the information they need. Process engineers at the plant are testing product-management software from Spotfire Inc. that offers visualization capabilities. TI is building custom templates to bring the data engineers need into the Spotfire tool. One template, for example, will be used to pull semiconductor defect data from a database and overlay it with electrical test data from another to determine which defects in the manufacturing process result in electrical faults.

Worthington uses a system called "inventory viewer" that taps into each division's enterprise resource planning system and collects raw material inventory data. Managers at all Worthington steel plants can view the information, which is stored in an Oracle data mart. That helps avoid situations in which one business unit buys raw steel while another has excess inventory.

Digital dashboards, often backed up by a centralized data warehouse, are becoming a popular way to get data to decision makers. In February, TruServe, a Chicago wholesale hardware company, launched a data warehouse and a dashboard, based on Business Objects SA software, for senior executives. Today, the warehouse contains information on stock sales and direct sales, franchisee claims data, and production data from TruServe's paint manufacturing facility. Soon, corporate financial data will be added to the mix. TruServe won't limit access to top managers; dashboards for second-tier execs such as distribution-center managers and inventory analysts are being tested.

"This is a tool we're going to put into every manager's hands, so we don't make decisions on gut feelings anymore," CIO Hastie says.

The system will lead to more timely decisions; managers will be able to learn on Monday how consumers responded to advertising flyers in Sunday newspapers, he says. Hastie expects the system to be especially helpful for making decisions relating to quality assurance and productivity.

But technology is only part of the fix. What's also needed is cooperation among IT and business managers. At Worthington, IT managers and business execs are jointly developing procedures for creating business forecasts and writing standards for divisions to submit consistent data to the data warehouse. "I think the biggest problem, from talking to my IT colleagues, is too much emphasis on technology and not enough on business," Dove says.

Dove meets with the heads of Worthington's business units to understand just what kind of information they expect to get from the data warehouse. Those requirements will set the direction for the technology the company will select to build the system. Oracle's database, a standard inside Worthington, will provide the platform. But the type of business-intelligence tools needed, and how much custom development Dove's team will have to do, will hinge on whether executives merely want to extract information from the data warehouse (which packaged tools can do) or add their own forecast and historical data for analysis, Dove says.

At Cigna Corp., an employee-benefits company with $85 billion in revenue, managers within the group insurance division make day-to-day decisions on managing the business based on financial and operational data, showing fluctuations in financial and operational metrics such as jumps in paid-out claims or slowdowns in claims processing, says Marc Bloom, the division's business technology and information services director.

Bloom's organization uses tools from Brio Software Inc. for ad hoc query and reporting. Some tech-savvy managers are able to build their own reports, others access the data using templates prepared by the IT department, and still others get a complete report delivered to their desktops. Matching the technology to the users is the challenge.

"Everybody's got a different skill set," Bloom says. "It's a question of having the right people and the right tools to access it, turning that data into useful information, and being able to do it quickly enough to make a timely decision."

Do managers at Cigna get the information they need when they need it? "I think they get more than half of what they need," Bloom says. For Bloom and other IT executives, the challenge is to get that number up to 100%.