Revving the Procurement Engine
A high-powered search tool steers Harris Corporation's engineers through immense volumes of data to locate top-value product components.
In the manufacturing supply chain, engineers wield a great deal of power. When an engineer specifies a component for a product, the company usually buys it. From a functional perspective, it's a solid choice. But from a business point of view, is it the best?
Maybe not. A different supplier might have another component that's just as good but less expensive. Or thousands of that alternative part might already be sitting in the manufacturer's warehouse. Maybe the engineer chose a part that the supplier has just stopped making—available now but impossible to find six months down the road. Or the supplier can't deliver for six weeks, while a competitor has inventory ready to ship.
Last year, engineers at Harris Corporation, a communications and information technology equipment manufacturer, gained the integrated view they needed to make more effective procurement decisions. The company worked with Endeca Technologies of Cambridge, Mass., to tailor an information search and navigation solution to its requirements.
At Harris Corporation, based in Melbourne, Fla., engineers who specify components have always had plenty of data available to help them make wise procurement decisions. In the past, though, the problem was how to make that data yield the best possible decisions.
"I had significant investment in managing internal data, but no integrated view," recalls Janice Lindsay, vice president, strategic sourcing at Harris.
Officials at Harris wanted to reduce component inventories and get new products to market faster. They wanted to avoid buying parts that would soon be obsolete. And, naturally, they wanted to save money.
Harris incurs nearly $1.25 billion a year in direct expenses—the cost of goods that it incorporates into products and services. In the past, engineers used a variety of tools to help them choose how to spend that money.
They conducted Google searches, relied on their collective "tribal knowledge," and consulted the company's Agile Product Lifecycle Management (PLM) solution from Oracle to see what components they had used in similar products in the past. But there was no systematic way to make those tools serve a single strategy.
"It was a hunt-and-peck system, and engineers had to know a little something before they could even get started," Lindsay says.
There was no guarantee that an engineer would choose a component from one of Harris's preferred suppliers, and no sure way to leverage the company's purchasing power across projects and divisions to gain better pricing. Nor was there any way to guarantee that the chosen part would be available quickly, or that it wouldn't become obsolete.
Enter Endeca's powerful search solution—and the beginning of a new outlook for Harris's engineers.
Like Shopping Online
Endeca's solutions help people steer quickly through information from multiple sources to reach the answers they need. One major application for this technology is to help buyers choose products, whether they're engineers at Harris Corp. or consumers shopping on sites such as HomeDepot.com or Walmart.com.
At the heart of any Endeca solution is a set of tools called the Endeca Information Access Platform (IAP), powered by MDEX Engine technology. The IAP processes data from multiple systems—both highly-structured data found in a database or table, and lessstructured data found in an e-mail or Word document.
The big difference between Endeca and a simple search engine lies in the way Endeca synthesizes and summarizes the data it finds.
"Endeca's advantage is its ability to provide an integrated view of the data," says John Andrews, director of industry marketing at Endeca.
Google and other traditional search engines are useful for searching the Web, but they don't apply enough criteria to return the results that engineers find most helpful.
For example, suppose a project requires one-inch-long stainless steel bolts with 16 to 18 threads per inch. "Typing 'bolts' into a search box yields one million results," Andrews says. "There's no way for the search engine to know what the user is actually looking for without more specifi c requirements."
Endeca's guided navigation technique quickly narrows the search to find products that meet all the criteria.
One way to get a sense of what Endeca does for engineers is to see what it does for consumers. Go to HomeDepot.com, enter "lighting" in the search box, and the search engine will return more than 4,000 results.
But use the menus to the left of the display to narrow that search by any of several criteria, such as category, price, brand, and "eco-option."
Navigate through those and subsequent choices that appear, and you quickly steer a path to, for example, five white ceiling fans and lighting fixtures, designed for small rooms, that cost between $100 and $200.
Searching High and Low
When Lindsay and her team started looking into ways to guide engineers toward better procurement decisions, they first considered standardizing on Google or a similar search engine. Engineers appreciate Google because it's fast, Lindsay says, but it isn't precise enough to meet their needs.
Harris chose Endeca after examining several search and navigation solutions. Endeca offered a proof-of-concept demonstration and showed that it could easily merge information from disparate sources into one application.
"Response time was fast and yielded relevant results. We could select custom views—per user or per category," Lindsay recalls. "That feature was unique to Endeca."
Harris signed an agreement with Endeca in 2006 and got the system running in 2007. Endeca already had a product designed to serve manufacturing engineers, but it didn't meet Endeca's needs right out of the box.
"Its basic design was leveragable, but we had to do a lot of custom work," Lindsay says.
Harris's engineering search application helps engineers make choices that support the company's supply chain goals. Implementing the solution resulted in cost savings, reduced component obsolescence, and shorter product development cycle times.
Harris has four major divisions, each with its own database of approved manufacturers and enterprise resource planning (ERP) system. All those sources provide information to the data warehouse that Harris and Endeca developed to feed the solution.
Harris's Endeca system also draws data from external sources. One important source is PartMiner, which enriches Harris's own parts data with information on factors such as technical attributes, obsolescence, product lifecycles, and regulatory compliance.
Another source is Dun and Bradstreet, which provides insight into suppliers' financial health.
Parts distributors share information on what products are in stock and how soon they can ship. Harris's cont ract manufacturers also contribute information.
"They enrich our data with the components that they're buying, so I can leverage their spend and my spend," Lindsay says.
Harris's engineers had an easy time learning the Endeca system, but adjusting to the cultural changes the system created was another matter.
"Procurement doesn't usually tell engineering where to go for sourcing information," Lindsay says.
To get engineers used to considering supply chain requirements, the procurement department developed case studies that showed how Endeca supports more effective decisions. Harris also introduced a policy requiring thatbefore the company releases a new product, the engineer responsible has to evaluate the bill of materials.
"They have to show they're not requiring obsolete products," Lindsay says. "Eighty percent of the parts have to come from preferred partners. If RoHS [restriction of the use of certain hazardous substances] is an issue, the part has to be RoHScompliant."
Using Endeca to procure parts helps engineers meet those mandates.
To make the solution even easier to use, Harris and Endeca currently are developing interfaces between that system and the software packages engineers use to design products.
Harris implemented Endeca as part of a larger program to transform its supply chain. The software is helping engineers make choices that support the goals of that initiative.
They no longer have to guess which of 10 components that fit the need on technical grounds is the least expensive. Also, thanks to Endeca, Harris is reducing component obsolescence and shortening product development cycle times.
"While the Endeca search capability is just one piece of the pie," Lindsay says, "it's the linchpin to our entire supply chain transformation."