What is the one thing without which a world-class supply chain cannot exist? Accurate data. But that's easier said than deployed.

From tracking goods until they arrive at a customer's door, to handling security and regulatory compliance issues, to optimizing inventoryeach step of the chain requires not only access to accurate data, but the ability to analyze the data to achieve optimal goods and information movement. At the same time, the volume of data that companies must master is increasing dramaticallyand companies need to stay on top of the data heap to be competitive.

We are in a period of data explosion that has been occurring for many years, says Paul Kirby, senior analyst, Forrester Research. That's been particularly true over the past three years as a number of factors have brought the data issue to a head: regulatory compliance requirements with such measures as Sarbanes-Oxley, supply chain optimization needs and continued integration of enterprise applications to improve business performance. Mergers and acquisitions, which require companies to consolidate internal data and systems from disparate sources, are another reason.

As a result, companies are turning to data warehouses to assist in collecting, securing and analyzing their data. Organizations implement data warehouses to bring data together from multiple sources to make better business decisions. Data is coming out of the silosilo applications are falling away and are more integrated, says Kirby. Without a data warehouse, data cannot be easily shared across an organization, creating pockets of information instead of a broad business view.

How does a data warehouse create a broad business view? By serving as a repository of transaction data specifically structured for querying and reportingand supporting management decision-making. It complements existing applications such as enterprise resource planning or supply chain planning system. For example, a data warehouse can add scalability to a transportation management system and provide one version of the truth across multiple systems. Data warehouses can house a wide variety of data to provide a snapshot of current business conditions. The technology includes systems to extract and manage data from operational systems.

In addition to uniting information from a variety of sources, data warehouses analyze data faster than an operational system. In most cases, companies can't run business intelligence reports directly on transaction processing systems, because they impact the performance of these systems, said Dan Vesset, research director, analytics & data warehousing, IDC.

Data warehouses can also handle information that other systems can't, i.e. unstructured data, which, according to Mark Beyer, research director, Gartner and Vesset, represents a whopping 80 percent of a company's data. Unstructured data includes, for example, specialized instructions on a bill of lading as well as documents, emails and multimedia. Traditionally, text instructions are not been picked up by an operational system.

Putting it in action

Kirby estimates that about 60 percent of companies with revenues over $1 billion are using a data warehouse in some form. General Motors and Union Pacific fall within that group. Each is using an NCR Teradata data warehouse technology to manage, store and dissimulate data effectively across geographies, systems and companies. Companies choose NCR Teradata's data warehouses to handle extremely large, multiterabyte amounts of data, says Lou Agosta, formerly with Forrester Research.

Wal-Mart, for example, is using NCR Teradata data warehouse technology to manage in excess of 560 terabytes of information (the equivalent of 38 miles of full file cabinets, 250 million pages of text) down to minute levels of detail, including items on the retail floor, supplier and carrier information and customer data.

Like GM, a large portion of Teradata's customers have international operations and use the technology across their enterprise, says Ed Dupee, solution vice president, demand and supply chain solutions. These international customers are using the data to track goods throughout the supply chain, predict issues, and prevent shortage.

Some of the virtues of storing data in one place are accessibility and error reduction. The same data populates multiple systems, reducing re-keying of information or incongruous data sets. By the same token, the data is more accessible, as it resides in a single place that can be accessed by multiple systems instead of in a system that may or may not be able to send data elsewhere. This information can then be distributed to different computer systems around the world and across trading partners.

Union Pacific exemplifies a company using its data warehouse to consolidate business operations. A railroad with revenues of $12.2 billion, UP installed the technology to get a consolidated, global view of its operations, streamline its supply chain and improve its customer service. The technology combines information from disparate systems under one umbrella, said Roger Bresnahan, principal engineer.

Before our data warehouse came into being we had stovepipe systems, he said. None of them talked to each other. We had anomalies that grew out of that. And we couldn't get a whole picture of the railroad.

UP began its data warehousing project in 1992 with its accounts payable information.

We built a foundation area with seven different subject areaswaybills, customers, trains, cars, car hire, accounts payable and accounts receivable, as well as equipment information, said Bresnahan. Twenty-six different departments were involved in the project.

Customers can access pricing information over the Internet, improving their ability to manage the financial end of their supply chain. Locomotive dwell time has been reduced by 39 percent, saving customers time, while asset utilization and crew management has improved by millions of dollars.

In addition, UP uses the data warehouse for forecasting, optimizing its supply chains by determining the best traffic corridors and most profitable market segments, where cars are needed seasonally and managing fuel costs.

The system paid for itself in less than three years, said Bresnahan.

Optimization tool

Data warehouses can also be used to drive process improvements in an organization. Vector SCM (Supply Chain Management), worldwide lead logistics provider for GM, is using data warehouse technology and tools to save GM millions in transportation costs.

Novi, Michigan-based Vector SCM is a joint venture between CNF Inc., a $4.2 billion transportation and global supply chain services company, and General Motors Corp.

Vector's data warehousing project seeks to optimize GM's inbound supply chain by improving GM's truck load utilization. The project started with inbound materials to GM's North American plants, said Paul Vollmer, director of engineering for Vector. GM has 65 plants in North America, 140 support facilities and 3,500 suppliers. In addition, some suppliers have multiple locations. Vector is optimizing transportation resources for 4,500 supplier locations and 300 carriers, involving 61,384 part numbers and over 91,417 supplier-part-plant combinations, he said.

Originally, Vector audited GM's inbound truckload network manually, said Vollmer. However, the process was time-consuming, subjective and somewhat dangerous. Now Vector uses NCR Teradata data warehouse software to automatically audit GM's loads. The system calculates load volumes and identifies areas of improvement in the network by tying advanced shipment notices to package data. Data is linked together in the data warehouse.

Now we can look at metrics: [volumes] by plant, route type, date, route number. We compare those numbers to the plan and tolerance levels, said Vollmer. Levels that are systematically outside tolerance are flagged and reported. Vector is using the information to look for long-term problems, not one-offs, he said.

Vollmer sees initial potential savings in the millions of dollars with the data that is currently available. Greater savings will occur when more data is entered into the system, he said.

Sidebar: Taking a Global View

Dusseldorf, Germany-based Metro AG, a $69-billion dollar retailer in 30 countries, is using its data warehouse as a one-stop global information center. Its data warehouse powers several applications, including its customer relationship management, load management, decision support and supply chain software, including its supplier portal, stock and shelf management and efficient consumer response tools, says Ralf Morgen, division manager, customer centric systems, MGI Metro Group. The retailing giant, which established in late September a radio frequency research and test lab with GS1 Germany, has linked selected data from its data warehouse to its RFID systems, said Morgen. MGI Metro Group is the company's information technology arm.

The aim of our data warehouse is to gather information and make it accessible for different purposes, said Morgen. Our data warehouse is now up and running in 18 countries. This year we will add five or six countries and more to follow in 2006. The data warehouse will support 28 countries, 24 languages, different character sets, formatting and cultures, he said.

Morgen cited two examples of how the company uses its data warehouse: as a customer relationship management tool and as a supply chain operations tool. Metro uses data from the data warehouse to help the company with promotions planning and execution, segment customer groups, and select marketing and direct mailing tools for a promotional campaign. The company can sort data by customer demographic, product and store location; or select stores by high or low sales.

Metro also developed a supplier portal that has been operational as of last year, said Morgen. The portal gives suppliers a tool for collaborating with Metro, including tools and applications, he said.

Without a data warehouse, Metro would be unable to easily collaborate across continents, merge customer data with operations data, or analyze customers by demographic. Comparing and analyzing store sales would be arduous. Data in different languages and currencies would difficult to compare in a meaningful way. By consolidating this information in a single area, Metro gains control over its data and is able to use it in a more meaningful way to make enterprise-wide and supply chain wide decisions.

Metro is like many retailers that are using data warehouses to make point-of-sale data available to manufacturers and suppliers and to enable all parties to optimize their supply chains, said Paul Kirby, senior analyst, Forrester Research. Data warehouse such as Metro's can be used to optimize inventory buying decisions, reduce inventories and make buying decisions based on actual, rather than anticipated, usage, he said.

Sidebar: An Array of Data Warehouse Offerings

Companies offering data warehousing in addition to TerraData include Oracle, Microsoft and IBM. The most notable is Oracle, with the largest market share. Indeed, many companies chose Oracle because they were already using Oracle for other applications, including enterprise resource planning.

Oracle's customer list includes online retailer Amazon.com and energy and petrochemical company Shell Group, part of Royal Dutch Petroleum Co. Both are using Oracle to create system-wide information viewsin Shell's case, to link geographic data from a variety of legacy systems and adopt geographic information system standards.

We selected Oracle9i Spatial because it supports GIS open standards and satisfies our requirement for interoperability. Oracle Spatial allows us to deploy geo-information systems in a multi-user environment, said Jerry Larthe de Langladure, head of Geoscience Leadership Team, Shell International Exploration and Production (SIEP).

SIEP, part of Shell's Exploration and Production organization, needed a central consolidated repository for information on all its assets, including oilfields, wells, and pipelines.

Oracle9i Spatial contributes to an integrated management system so pipeline engineers aren't just concerned with pipelines, said Larthe de Langladure. They are able to take into account the fields and roads they will use so they can consider all the environmentally sensitive issues in areas they might want to put a pipeline. It gives a more complete picture of the pipeline environment. Drillers not only know how to get from the surface down to the target environment, they gain access to information to determine where to put the rig so it will be unobtrusive to the local population.

This information was not readily accessible without a data warehouse, said Larthe de Langladure. The inability to consolidate proprietary and third-party data prevented us from quickly getting information to the decision makers that execute our global strategies, he said.

Online retailer Amazon.com, with net sales of $7 billion, is using Oracle to improve its customer service and hone its one-to-one marketing initiative. Amazon chose an Oracle data warehouse for a single view of the business, improved site performance, flexibility and scalability, said Mark Dunlap, director of data warehousing for Amazon.com.

By using a data warehouse, Amazon.com can understand and manage its entire global business at any level, from enhancing customers' experiences while visiting Amazon's web sites to managing product flow through the entire supply chain to providing business intelligence for Amazon.com management.