Welcome warehouses.

Information is the lifeblood of companies. Without timely, accurate information about customer buying trends, operational performance, and marketing campaigns, business grinds to a halt. Unfortunately, in most companies, this information is either imprisoned in arcane data base structures or randomly disseminated throughout the corporation via spreadsheets, PC data bases, and functional or departmental applications.

The solution is a data warehouse, which consolidates and stores historical data drawn from operational data bases, integrated along common business dimensions, such as customers, products, and regions. Thus, users can directly obtain and refine data from different software applications without affecting the operational data bases. They also can integrate different business tasks into a single, streamlined process supported by real-time information. Data warehouses run on Unix and Microsoft/NT client/servers at a fraction of the cost of a mainframe computer. But as companies such as State Street Bank and Merck have discovered, the real payoff comes not necessarily in dollars, but rather in time saved, better decision making, and improved corporatewide communication.

The data warehousing/executive information systems project at Boston-based State Street Bank was initiated by the company's chief financial officer and implemented early last year by one of the bank's senior financial managers. Such high-level involvement and sponsorship have been key to the initiative's success. In about nine months, a team of Information Systems professionals and two software companies created, a point-and-click EIS and a comprehensive data warehouse that organized and refined the data from a variety of sources, including the company's general ledger system, profit center accounting system, cost allocation system, funds transfer system, demand deposit data base, and an asset management application.

As a result, more than 150 executives now have access to up-to-date, easy-to-use information about the bank's many operations, helping them to make better and faster decisions. They can slice and dice information across time (year, quarter, month), by scenarios (this year versus last year, actual versus budget, forecast), within and across legal entities and profit centers. For example, now they can instantly compare a graph showing a dip in net interest revenue in one of their operations with changes in business activity or deposit levels in that division. Being able to correlate such cause-and-effect relationships enables the bank executives to make policy decisions or to launch targeted promotions to bolster sagging earnings.

Data warehouses often are a byproduct of integrating applications necessary to complete a single business process. In identifying and capturing key events - such as generating a purchase order or moving inventory - that make up each business process as they occur, a dynamic information repository is created. This data warehouse allows applications to share important and relevant information.

Ian Miller, executive director, Automation and Information Technology, at White House Station, NJ-based Merck and Co., used this approach to integrate Merck's procurement, accounts payable, inventory management, and production systems into a single procurement process. In the increasingly competitive pharmaceutical environment, Merck is placing a much higher value on the information it has about its business. Managing information effectively can make the difference between having three and eight months of inventory sitting on the shelf - a considerable savings. Since each of the applications that needed to be tied together had been built on a different system, Miller's team decided to have the applications simply pass the important business events to one another.

For example, the procurement process comprises five events: Create a purchase order. Approve a requisition. Move inventory. Approve a batch. Release a lot. Every time one of these events occurs, the application involved "publishes" the relevant information about it. Any other application in the business process that needs to know about it "subscribes" to the event. The data warehouse becomes one of the applications that "subscribes" to all significant business events, creating a two-way flow of up-to-the-minute business-process information.

By focusing your Information Technology staff on integrating applications using business events, you'll gain a streamlined, integrated business process and a dynamically updated warehouse of important decision-making information.

Patricia B. Seybold is president and chief executive of the Patricia Seybold Group, a Boston-based information-services firm.