Why Wesco Doesn`t Use ERPBy Mel Duvall | Posted 2007-05-14 Email Print
Re-Thinking HR: What Every CIO Needs to Know About Tomorrow's Workforce
Wesco International didn't want to pony up big bucks for traditional enterprise resource planning system. How a new, quicker data warehouse helped the $5.3 billion company anchor its business.
As a Fortune 500 company with $5.3 billion in revenue in 2006, Pittsburgh-based Wesco International plays with the big boys. The electrical and industrial product distributor has more than 6,000 employees, 370 full-service branches across the U.S. and Canada, eight high-tech distribution centers, and more than 100,000 customers worldwide.
What it doesn't have is an enterprise resource planning (ERP) system.
Over the past five years, Wesco has undertaken a major overhaul of its computer systems, giving its managers closer to real-time access to sales and inventory information from the field. It has added business intelligence capabilities, allowing sales analysts to quickly determine exactly which customers are buying more or fewer products. And it has added applications to automate more human-resources functions, such as annual employee evaluations and calculating merit pay.
But unlike most corporations of its size, which have added these capabilities by installing large ERP systems from Oracle, SAP or similar vendors, Wesco chose to take a radically different approach. It has created a de facto ERP system by building applications that tap into a Teradata data warehouse, which has become the centralized repository for most of the company's business information.
It's an approach even Teradata says it wouldn't necessarily recommend for most companies. "We don't actively pursue a Teradata data warehouse as a replacement for an ERP system," says Shaun Connolly, a senior consultant in Teradata's transportation and logistics industry practice. Most companies find that the automated workflow features that ERP systems offer, such as the ability to route sales or purchases for approval, are often difficult to duplicate, according to Connolly. "I think people are surprised more than anything that a company the size of Wesco has chosen to go down this path," he says.
To understand why Wesco chose this approach, you have to first understand its business. The company dates back to 1922 when it began as the distribution arm of Westinghouse Electric, selling Westinghouse-manufactured products across the country. Its strategy has centered on putting inventory, expertise and services where its customers need them. Wesco's customers cross most industries and run the gamut from Boeing to Dow Chemical to PepsiCo. As a result, it has a highly distributed organization with 370 branches, fed by eight distribution centers.
Distribution center managers are given a high degree of autonomy, including the ability to determine inventory, set prices and negotiate contracts. That strategy served the company well from a sales and operations standpoint, but corporate visibility into branch activities was limited.
The Pittsburgh head office, for example, did not have real-time access to inventory at the branch level, and could not as a result easily shift supplies from one location to another to meet demand. It also did not have immediate access to sales information from the field; this data was consolidated at headquarters via nightly uploads to an Informix database. While management could then gain access to top-level numbers, such as overall sales, it could not quickly drill down to important customer-level information, such as which customers had recorded a dramatic drop in purchases and were perhaps getting their supplies from a competitor.
The Informix system, which was installed in 1993, couldn't be tweaked much further. John Conte, Wesco's chief information officer, says it was overloaded and underpowered.
A key sales analysis report, which formatted a full month's sales transaction data for analysis by managers, required 80 hours of processing time an eternity by today's standards. "It's one of our most CPU-intensive reports," Conte says. "It rolls up a tremendous amount of data by region, inventory levels and pricing information, and packages it so it's actionable by the field offices."
Wesco knew it needed to upgrade the system to bring it closer to real time, but the economic downturn of 2000-2001 brought the message home. In 2000, Conte, who was I.T. director at the time, and then-CIO (now chief financial officer) Steve Van Oss, began evaluating an ERP system to move closer to real-time visibility into Wesco's operations. The company looked at systems from SAP and Oracle, among others, but the choices were not very palatable. For starters, there was the cost. Conte says that to get the functionality Wesco needed would have cost close to $110 million.
In addition, to achieve the integration required to get real-time data from the field, Wesco would have had to scrap WesNet, its distributed point-of-sale system. WesNet runs on local servers at all Wesco branches and distribution centers, and is based on a 20-year-old NCR system called ITEM. This option was not at all popular with branch managers who were comfortable with WesNet, nor was it comfortable for Van Oss. WesNet was completely paid for, incorporated a high degree of customization, and could still be expanded.
Conte says it might have been possible to integrate WesNet into an ERP system, but in many ways WesNet provided the sales and inventory functionality that an ERP system was meant to provide.