Four Strategies for a Web-Based Supply Chain

More than one strategy is needed to create and manage your online supply chain.

"Whatcha gonna do now?"

For most of the Fortune 1000, the bully posing that classic challenge is the Internet. The Internet is not just about buying books through a website anymore; the Internet can handle your entire supply chain. Do nothing, stand pat, and you risk seeing your competitors lure away your customers with Internet-based supply chains that are faster and cheaper, and can offer customers more information about their orders than you could ever dream about giving them in a memo or over the phone. Accept the Internet challenge, revamp your business processes, share your company secrets with customers, suppliers and even competitors, and you risk throwing your business into turmoil.

It's not really a choice. Right now, as supply chain technologies and business models evolve and customer tolerance for failure is high, is the time to take risks.

Some clarity has emerged from the cloud of hype that is the B2B revolution. Some universal strategies have been discovered for constructing a virtual supply chain, a rough outline for the future of business. The clear message so far is that you won't be relying on a single online strategy. The Internet supply chain will be a variety of means of communicating and doing business with suppliers and customers. Below, we outline four primary options.

1. Online procurement—your introduction to a Web-based supply chain

You need a chair, and you need it now. But no one in your department knows where their chairs came from, and nobody is quite sure who's in charge of buying them. And, anyway, what's the big deal? You go to the nearest office furniture store, pick out the one with the leather cushions (you've earned it, right?) and expense it.

You've just committed the cardinal sin of procurement: the maverick spend. The maverick spend makes it difficult for procurement departments to track costs. It makes it impossible to budget accurately. It turns forecasting into guesswork and ultimately destroys careers. If you're running a business, it's a very bad thing.

Jeff Campbell, vice president and chief sourcing officer at the Burlington Northern Santa Fe (BNSF) railroad company in Forth Worth, Texas, knows all about the maverick spend and has stamped it out. Office chairs, computers, pens, pencils and the like have stopped being random purchases. Campbell knows what BNSF wants in a chair, whether it be for a lounge (no swivel or tilt) or behind a desk (don't expect leather unless you're really special). He knows from whom you should be buying it, regardless of which state—of the 23 BNSF serves—you work in. He knows what it will cost. He knows all this because the only place BNSF employees can purchase these things is on the Web.

It used to be that, like politics, all procurement was local. "Even if we made contracts with suppliers, people still went out and bought from local mom-and-pop stores," says Campbell. That meant that Campbell could not simply automate BNSF's old procurement process and put it on a website; first he had to centralize purchasing across all of BNSF for even the most basic materials.

In order to make himself the final authority on purchasing, he had to drag the maligned and ignored job of procurement out of the depths of his company and up to the strategic level. With his extensive experience in procurement at other companies, he was able to swing a direct reporting relationship to the CEO when he came to BNSF. "At most companies, purchasing is viewed as tactical. But Campbell was elevated to an important role. It made a big difference to the project," says Steve Hornyak, chief strategy officer at Suwanee, Ga.-based Clarus Corp., which makes the procurement software BNSF uses.

BNSF's push for Internet-based indirect procurement—items like pencils, pens, chairs and computers—is a good first step toward a full-fledged electronic supply chain because it does not affect the day-to-day business of the company, just the administrative side. Though the payback for a typical e-procurement project is not as great as for other, more far-reaching electronic supply chain projects, the savings can be significant, and the time to implement the projects can be mercifully short. Even better, a successful procurement project can act as proof-of-concept for e-commerce skeptics in the company.

Best of all, procurement projects involve the same strategies as fully Web-enabled supply chains but on a much smaller, more controllable level. Campbell began by examining the specifications for all the different products BNSF uses, from mops to mouse pads. The group then decided whether the products were sub par, just right or "overspecced" (did the chairs really need lumbar support levers?).

Next came the task of trimming the supply base—the biggest potential cost savings for BNSF if done well. Campbell had to find suppliers that could get products to all the states in which BNSF operates. Then he had to figure out what the best price really was, factoring in shipping costs, warehouse costs, penalties for late delivery and the like. And to make sure BNSF is paying the actual cost agreed to in the supplier contract, Campbell created a set of metrics to keep an eye on supplier performance. Much like Staples' supplier website (see "The Big Payoff"), the BNSF site will push metrics to the suppliers so they can see how they are performing any time they want.

So far, Campbell estimates savings of between 3 percent and 28 percent on the different indirect purchases made through the employee purchasing site, called BNSF Source Net. "The volume of purchases is so much larger now that we can negotiate incremental discounts on just about everything," he says. There are internal savings as well. Approvals for purchases, normally a time-sapping, manual process, are now automated with forms online.

Campbell can also keep an eye on those mavericks out there. Everyone at BNSF is required to use Source Net. If the item they want isn't on the site, they can describe it to different suppliers online to see if one of them can provide it. If all else fails, employees can use their corporate credit card to expense it. Just don't try to hide it. The system yanks indirect procurement items from expense reports and presents them as part of the overall procurement reporting system, by commodity type, vendor, department and user. (Um, wanna explain why you expensed that Palm Pilot—the one with the color screen?)

2. The consortium—speaking the same language

Guy Abramo's company, Ingram Micro, a Santa Ana, Calif.-based wholesaler of technology products and services, works with computer manufacturers and resellers that create 250,000 products and parts of products. Of course, no two suppliers call any one of those 250,000 items by the same name, and no two suppliers' invoices or order forms look the same. One's X-19 transistor is another's Y-18. It's the same transistor, but who's to know? Naturally, this served to dampen CIO Abramo's enthusiasm for electronic linking.

IT's traditional method for resolving nomenclature differences is to create a translation technology—sort of like headphones at the United Nations—a middleware, so that the different computer systems know what the other is talking about. But Ingram Micro is part of the 2-year-old RosettaNet high-technology consortium of companies that is tackling the problem at its source. The 250 members of RosettaNet are trying to create a new universal language, a kind of high-tech Esperanto.

RosettaNet standards are constructed to be simple, yet comprehensive. The new language is expressed in Partner Interface Processes (PIPs), which define a particular business process. Now, when CompUSA wants to find the price and inventory status of a particular PC from Ingram Micro, it sends an XML document that has been written to the PIP standard. That message dives into Ingram's mainframe and instantly relays back the price and availability of the product. The PIPs squeeze out the demon at the core of computer-to-computer integration—variability—by leaving nothing open to interpretation, from the ways to perform a process and talk about it to the technology for exchanging information about it.

Needless to say, the workings of this Geneva convention of supply chain modeling have been neither easy nor quick. Getting companies to swallow their pride and admit that another might have a better way of constructing an invoice, for example, is difficult, especially when 250 companies are trying to do it at once. Two years into the consortium, RosettaNet has just 10 PIPs on its platter, and Ingram Micro has implemented just two of them: price and availability. But if RosettaNet takes off—which seems likely given all the industry heavyweights (including Cisco, Dell and IBM) that have signed up—then Abramo's work, at least as it concerns looking up price and availability on his system, will be done. His customers and suppliers can get into Ingram's system with very little intervention from Abramo and his staff. "We do 60 million transactions a day with a host of different systems," he says. "With the standards, if I do the integration work for a process, I can do it once and be done with everyone. As long as the company connecting in uses the RosettaNet standard, I don't have any work on my end."

But Abramo's work doesn't begin and end with system-to-system hookups and lookups. Right now, Ingram Micro spends a lot of time and man power manually reconciling part numbers across manufacturers and digitizing product specifications, descriptions and photos for placement on its website. If PIPs were created for exchanging part and product information, Abramo could flow all that information directly into the website, eliminating all the faxing, phoning, retyping and image tweaking that goes on now—tasks that delay the introduction of a new product into Ingram's catalog.

"With the PIPs, as soon as IBM announces a new product, I would be ready to accept it into my warehouse," says Abramo. Considering the surreal speed of new product introductions in the PC industry, the "build it once and they will come" approach of RosettaNet could save Ingram "tens of millions of dollars yearly," Abramo says.

The PIPs would also light the remaining dark tunnels of the high-technology supply chain—like the small retailers who, according to Abramo, handle 80 percent of the PC demand in the United States. "We already have a highly automated supply chain with our large customers and manufacturers; it's the small and midsize companies that need the standards. We've never been able to afford to hook up with them on a onesey-twosey basis, and they couldn't either." With PIPs, Ingram will be able to replace a major step—and expense—in the supply chain for smaller retailers.

Today Ingram Micro ships 72 percent of its products directly to end customers, thereby keeping the products from lingering needlessly in the retailer's warehouses. With smaller stores hooked into Ingram through the PIPs, they too could get in on that action. "We have to ship it to them, and they stick it in their van and get it to where it needs to go. We could be shipping directly to their customers," says Abramo. Instead of passing product, Ingram Micro and the retailer could be exchanging information about the order—quantity, shipping cost, address and the customer's expected delivery date. "It becomes up to us to manage the information flow that has taken that step out [of the supply chain]," says Abramo.

PIPs could also light the tunnel that connects the PC manufacturer with end users. One of the reasons that PC maker Dell, of Round Rock, Texas, has thrived is because it sells directly to its customers; the company has real-time feedback about what its end customers want and don't want. For the IBMs and Compaqs of the world, the lights go out at the distributor level. "If we can get IBM real-time information on how its products are moving through our channel, it will bring light-years of improvement in demand planning for IBM," says Abramo. "It's not that we don't want to share those numbers with IBM right now, it's just that our systems aren't set up to do it." A PIP that enabled Abramo to build that kind of tracking capability for all his suppliers would make it worth everyone's while.

Though critics complain that RosettaNet, like most standards bodies, is slow and that its process-by-committee strategy will compromise the effectiveness of the PIPs, that hasn't stopped similar organizations from sprouting up in other industries where margins are low, product turnover is high and consumers demand enormous ranges of choice—like consumer packaged goods, for example, provided by companies like Procter & Gamble, Wal-Mart and the major grocery chains. There, a nonprofit group, the Voluntary Interindustry Commerce Standards Association, has created a set of standard business processes for the retail industry known as Collaborative Planning, Forecasting and Replenishment.

"A year ago, I would have said that high technology is the exception to the rule in terms of being able to collaborate on standards, but now we're seeing it in other industries too," says Abramo. "I know the automakers have approached RosettaNet to see how it does it and whether PIPs might have applicability in their industry."

3. The private portal—the customer is always right

1 2 Page 1
Page 1 of 2
Download CIO's Roadmap Report: Data and analytics at scale