An IT Fixer-UpperAn IT Fixer-Upper

Home Depot Inc. is doing some home improvement of its own. The $54 billion remodeling-products retailer is overhauling its outdated decision-support systems in an effort to reap greater profits from an ever-growing number of stores, stand up to tech-savvy competitors' aggressive expansion efforts, and boost its languishing stock price. It's a big job for Home Depot's business-technology management team, involving the construction of a huge data warehouse for analyzing sales and operations, a test of self-checkout systems at some stores, and the deployment of new financial and supply-chain management software.

The data warehouse project unveiled last week is the most visible sign of change initiated by CIO Bob DeRodes, who took the reins of the IT department at the nation's second-largest retailer in February. The initiatives show that DeRodes has the backing of CEO and president Robert Nardelli, a former General Electric Co. executive who assumed his post 11 months ago. At a time when IT spending at many companies is sharply lower, Nardelli signed off on a project with a price tag in the tens of millions of dollars. Home Depot won't go into details about its tech budget, but Nardelli has stated that he intends to increase IT spending to accommodate the company's growth.

Change is needed. Home Depot has been a technology laggard in some areas, says Paula Rosenblum, AMR Research's retail research director. Business intelligence is one of them. Asked to assess Home Depot's IT situation, DeRodes says: "We're not as well-positioned as we'd like to be," referring to the company's computer environment as being "classic legacy."

With some 1,500 stores around the country and a new store opening every other day on average, the limitations of Home Depot's existing infrastructure are increasingly apparent. To gain insight into sales, pricing, and other areas, managers must ask programmers to tap into as many as 16 mainframe-based operational systems, extract the data, and create reports for analysis. That's a lot of work for the IT department, but more important, it's too slow for good decision-making. "We do that with great pain and angst," DeRodes says.

Without a data warehouse, quick and accurate decisions about what products to sell in which stores at what price are impossible, Jupiter Research senior analyst David Schatsky says. "That's the opportunity for them, making optimal choices about those things."

No. 2 home-improvement retailer Lowe's Companies Inc. already has embraced that opportunity. Lowe's is building new stores almost as quickly as Home Depot is-sometimes in the same neighborhoods. Lowe's often wins when the two retailers go head-to-head, Rosenblum says, "and IT is part of the reason." Lowe's, which is less than half the size of Home Depot, with $22.1 billion in sales last year, uses a data warehouse and analysis tools to track customer-buying patterns and tailor promotions in different regions.

DeRodes is confident Home Depot will catch up. "It's pretty unusual for a company our size to be just starting something like this," he says. "But it becomes an advantage. We're not going to waste a lot of time learning." Data warehousing technology and practices have matured, he notes, and IT workers with experience are plentiful.

DeRodes himself is no novice with such systems. He served as CIO of Delta Air Lines Inc. and CEO of its technology subsidiary for nearly three years before moving to Home Depot. The airline has long been recognized as a data warehouse pioneer. Managing a data warehouse was also a big part of DeRodes' job at Citibank's global cards product division, where he served as chief technology officer in the mid-1990s. In June, DeRodes appointed Kevin Murphy, who helped him build Citibank's data repository, to the key post of information management VP in charge of overseeing the business-intelligence project.

Home Depot will launch the data warehouse next month with applications for analyzing human-resource expenses. Built on IBM's DB2 database, the system will run on two 32-node pSeries 690 IBM servers with the AIX 5.1 operating system and 60 terabytes of storage. The company is negotiating with vendors for data-integration, reporting, and statistical-analysis software, Murphy says.

Starting early next year, point-of-sale data will be loaded into the system for a broad range of sales forecasting, product pricing and assortment, store-space planning, inventory management, and purchasing data-analysis applications. Initially, the system will contain 5 to 10 terabytes of raw data.

The data warehouse will be available to top executives, midlevel and store managers, and business analysts, with tools tailored for each, Murphy says. That will help managers identify and keep in stock the most profitable items. The warehouse also will provide a continuous flow of business-performance data in near-real time to executives and managers using management dashboards, letting them identify and respond to, say, inventory bottlenecks. Nardelli knows the value: Dashboards are in wide use at GE, where Nardelli was CEO of GE Power Systems.

Home Depot's IT systems are getting more than a face-lift. At 60 terabytes, it's more like a forklift. But if DeRodes is successful, Home Depot will maintain its legions of loyal contractors and home-improvement enthusiasts.