Three years ago, when North American steel service center operator Marmon/Keystone LLC contracted with Ruan Transportation Corp. to operate Marmon's truck delivery operations, the goal was cost savings and better service in shipping metal tube, pipe and bar to customers.
“Logistics has become so complicated, with all the reports you have to file to satisfy the government, and getting drivers is very tough,” says Norman Gottschalk, Jr., Marmon/Keystone president. “Why not get somebody who has all of this automated to start with?” What's more, “they have a pool of drivers; so if one quits we get one the next day.”
The numbers seemed to add up, but a potential “deal breaker” in the agreement had nothing to do with arithmetic or hiring. Signs on the trucks mattered, too. “If we couldn't have our own name, we weren't going to do it,” Gottschalk says.
Two things stand out in this familiar tale of business outsourcing and private labeling. First, a more than 100-year-old service center operating on longstanding principles of minimizing risk and maintaining control just recently gave up its own truck fleet and reached out to a third-party logistics provider, known in supply chain management jargon as a “3PL.” Second, the sweeping terms of the agreement, spanning logistic algorithms all the way to such basics as truck signage, explain much about the gap between theory and reality as the metals industry confronts the wide world of supply chain management. The issues range from the sophisticated to the elementary, from modern theory and computerized practice to where to hang your shingle. Best practices appeal to and work for some. Yet for many others in the metals value chain, best practices barely create much more than small ripples on the long-established operational fabric of the industry.
At either end of the metals industry supply chain, the giants in the industry—global metal mills and large-scale original equipment manufacturers (OEMs)—are hunting for ways to shave costs from the metal they produce and consume.
In a 2007 speech, Lakshmi Mittal, chairman and CEO of steel giant ArcelorMittal, declared the “age of discontinuity” in the steel industry was over and a “new continuum” was underway.
The precipitous slide in global manufacturing in 2009 postponed Mittal's forecast. The task, as Mittal and others see it, is to assemble a linked group of specialized companies, including miners, inventory managers, metal processors, transportation providers and final assembly operations—sharing data visible to all through the Internet, and all operating from the same playbook. Advances in digital communication, visible via online password, make the dream possible.
Here's the reality: The potentially seamless system, especially in North America, often breaks down in the middle of the chain, where independent metal distributors and processors conduct daily business in a tradition-steeped way based on relatively narrow supplier and customer bases. This segment of the chain sees value in the personal touch and continues to peg its fortunes to customers who expect low prices and timely delivery above all. For a number of reasons, the concept of a smooth electronic information flow envisioned by giants like Mittal doesn't, and in some cases can't, work in the face of efficient practices that have evolved to meet local business needs. “All steel is local,” says Bert Tenenbaum, chief executive officer of Chatham Steel Corp., Savannah, Georgia, a unit of Reliance Steel & Aluminum Co., stealing the famous line about politics voiced by the late U.S. House Speaker Tip O'Neill.
Mittal's global vision may yet come to pass. But a look at the metals supply chain reveals the problems as well as the promises of logistics modernity.
In a 2008 research paper, two analysts of industrial logistics, writing from Finland, found that aside from the volatility of iron, coke and coal prices, the biggest obstacle to profitability in steel production is “the lack of visibility across the entire steel supply chain. Despite having a well-recognized enterprise resource planning (ERP) software package, most of the steel companies have been focusing on new solutions for their supply chain.”
Historically, steel mills focused almost exclusively on maximizing production to make the best use of their massive investments in steel-making capacity. Mills were driven less by individual customer demand than by their need for optimal output, says David Sinclair, a Boulder, Colorado-based metals industry consultant and a former information technology vice president at steel bar maker Niagara LaSalle Corp.
“Their fixed costs are huge, and once they get to that incremental point [of output], they are like the U.S. mint, because their variable costs are so small,” he says. “As the economy slows, their incentive is almost to give the product away.”
Gordon Gustafson, a metals industry veteran and consultant with iSG2 Technologies, Inc., agrees. “Mills have an economic order quantity, known as a heat. How many tons of steel do they make with the same chemistry?” That amount is almost always greater than the logical consumption requirement of even the mill's biggest goal remains a work in progress.
OEMs “aren't competing against companies any more; supply chains are competing against supply chains,” says Trevor Stansbury, founder and president of Supply Dynamics in Loveland, Ohio, a unit of O'Neal Steel, Inc., and a vendor of technology that enables OEMs to control better their raw material and parts procurement through Internet-based communication.
The drive by major manufacturers to outsource parts production is an old story. What's different in the metals supply chain is the effort by OEMs to continually audit and control what their outsourcing partners are up to, rather than merely negotiate prices and delivery times for discrete orders.
Service centers “are being driven by their customers,” says David Closs, Ph.D., chair of the Department of Supply Chain Management at Michigan State University. “There is increased interest by durable manufacturing industries to find value-added in working with their suppliers.”
Auto maker Mitsubishi Motors North America, Inc. in Bloomington/Normal, Illinois, acquires parts for its assembly operations from metal stampers, which in turn obtain metal from service centers. But, as other major manufacturers do, Mitsubishi—not its stampers—negotiates the price of its metal before the stampers receive it.
“If we were going to direct the purchase of steel to the stampers, we wanted to have a better means of managing what they were ordering and at what cost, so we would have a better idea of what's being used,” says Greg Neville, senior buyer for Mitsubishi in Bloomington/Normal.
Stampers must place their orders for metal through an Internet-based system licensed to Mitsubishi. “We can track every order by any supplier in the program, and at any one time we've got visibility into what our stampers are ordering and when they are ordering it,” Neville says. Mitsubishi monitors how much its service centers and stampers purchase through a password-protected Internet platform. In turn, Mitsubishi's metal suppliers have access to its order book.
The system, initially marketed by ADS Logistics Co., LLC under the name LoMaS (Logistics Management System), enables participants in the supply chain to know how much weight of each type of metal is needed to meet Mitsubishi's production forecast going out 14 weeks. The system delivers nearly 60 types of metallurgical data and specifications about each order.
“You want the supply chain managed and you don't want any overstock anywhere,” says industry consultant Gustafson, a former ADS Logistics executive and a developer of the LoMaS system used by Mitsubishi. “The consumer manages the supply chain.”
The system even exposes details of scrap generated by stampers and enables Mitsubishi engineers to maximize the blanking process. “We're scraping millimeters off the width of things,” Neville says. “We can estimate how much waste the stampers are going to produce in a month.” If scrap metal prices exceed a pre-determined level, Mitsubishi shares in the windfall.
The Mitsubishi story is part of an industrial trend. In June, Caterpillar Inc. announced that its Progress Rail Services subsidiary agreed to acquire Electro-Motive Diesel from two private equity firms. Electro-Motive, based in La Grange, Illinois, makes diesel-electric locomotives. In the acquisition, Caterpillar will bring under its tent the Supply Dynamics supply chain management system called Oasis, which is licensed to Electro-Motive.
What's going on is an effort by OEMs to correct an error they made over many years in outsourcing their parts supplies to independent companies—namely the loss of visibility into raw market demand—by increasing the visibility of forecast inputs for suppliers as well as themselves, says Supply Dynamics' president Stansbury. OEMs want to choreograph their supply chain by boosting the visibility of demand-related data from the parts suppliers that comprise the extended supply chain, he says.
“It's not just a question of cost and on-time delivery. OEMs have recognized that by better understanding the bills of materials associated with the parts that they buy from outside suppliers, they are able to do things they were never able to do before, including communicate vital demand-related information to mills and distributors that allow those mills and distributors to operate more efficiently, instead of speculating on what demand is going to be. Then, the supply chain is much more responsive to changes in OEM finished part demand.”
This year, in a possible sign of things to come, steelmakers Nucor Corp. and Mitsui & Co. (USA), Inc., formed a joint venture to operate steel processor Steel Technologies, Inc., based in Louisville, Kentucky. The “NuMit” corporate consolidation is just one of many possible futures in the relations among metals supply chain links.
Only a small fraction of O'Neal Steel's service center sales participate in the Supply Dynamics demand aggregation program, says O'Neal chairman Craft O'Neal. “Not everyone is sold on it yet, but it seems to be growing.”
Still, advanced information technology has come to the service center business in fits and starts. “We went [to the service center market] with ERP, and everybody said, what's that? ERP means nothing to a service center that's been in business for 80 years,” says Scott Cranford, vice president for sales at Verticent Inc., Tampa, Florida, one of the ERP software vendors that sells to service centers.
Indeed, “a lot of service center people would argue that calling it 'supply chain' is just another name for the business they do every day,” says Bill Edwards, a metals industry veteran and independent business consultant in Libertyville, Illinois. “I'm suspicious that some purchasing guy in some industry invented this supply chain terminology just to elevate his status, but that's not really fair.”
“We've been preaching this stuff for 20 years,” says Roy Shapiro, professor of business administration at Harvard Business School and author of “Halloran Metals,” a fictionalized case study of business strategies at two service centers. The metals industry has advanced in adopting information technology and supply chain theory but is well behind the curve, Shapiro says.
Publicly traded service center companies promote their supply chain management skills to Wall Street but often decline to provide details. In a typical response to an inquiry posed to several service centers about their latest supply chain management innovations, Lourenco Goncalves, chairman and chief executive officer of Metals USA, Fort Lauderdale, Florida, says, “I have no intention to discuss that, because this is proprietary and I don't want anybody learning what we're doing.”
Worthington Industries, a metals processor based in Columbus, Ohio, has employed ERP systems since the mid-1990s. In a recent cost-cutting drive, additional software from Oracle has been installed. But cost/benefit details aren't being disclosed.
“It's a sharp focus of ours,” Worthington Chairman and CEO John McConnell told a quarterly earnings conference call in July. “Beyond that, while I shouldn't be so naÃ¯ve to think that none of our competitors understand what we're doing, I'd rather just leave it as a bucket of tools that we use and will continue to expand.”
In addition to guarding proprietary information, one impediment to supply chain management bragging rights is the fact that assessing the bottom line impact is difficult. In the last 15 years, as Worthington has embraced ERP and other IT solutions, its inventory turns have shifted with the economy, averaging seven times per year, with no trend in that number that could be linked to greater efficiency.
“You see the investments in (ERP) systems and they seem to be quite notable,” says equities analyst Luke Folta, who studies Worthington Industries and other metals companies at Longbow Research in Independence, Ohio.
“They are expensive systems. They seem to be hot topics at the time of the investment, but I don't know that it's highly tracked. To the extent that a company can improve its inventory turns and working capital management, that's where the benefits are. But to the extent of trying to bring it back to an ERP system, that isn't something I've done.”
“I'm not sure the market has distinguished between whether you have an Oracle system, an SAP system, or something that your grandfather handed down to you on a piece of scrap paper,” says Mark Parr at Keybanc Capital Markets, Inc., in Cleveland, Ohio, another metals equity analyst. “There are potential benefits to improving coordination across relationships, but there are competitive issues and issues of proprietary skills.”
Advanced ERP systems may have their biggest benefit in the service center arena by enabling acquiring companies to more readily digest acquisitions, Parr adds. “You're starting to see the mills get much more aggressive and focused on acquisitions moving downstream into the service center sector. It's being done around the world, with the U.S. being the exception.”
From an operational standpoint, however, service centers have not always adapted well to ERP systems. Shapiro's case study described two metal service centers in the same region of the country selling largely the same product line. One was a highly centralized organization, centered at a single, 400,000-square-foot warehouse and focused, compared to its competitor, on a relatively short list of high-volume customers, 3,000 accounts. The competitor operated in a decentralized structure, where managers of its four branches were “kings,” with power to select what products to stock and how to satisfy customers. The customer base was 10,000 accounts.
“The case is based on a real story,” Shaprio says. “It has to do with supply chain strategy. The two companies have very different strategies. There are many ways to skin the cat. One of them is highly centralized and highly focused on volume. Its strategy is to be low cost.” The other company “is very decentralized and very entrepreneurial. Branch managers rule over a small fiefdom of their own. It is very flexible and higher cost, so they can focus on service.”
Both strategies can work profitably, Shapiro says. But the appropriate supply chain management concepts and tools, especially regarding collaboration among mills, service centers and customers, probably will be quite different between the two companies, he says.
“There are things they can learn from one another, but the temptation to adopt elements of each other's systems is great but generally the wrong idea. Most students who don't know much about the industry say steel is steel is steel; there has to be one right model. It turns out that is wrong. Many things that the other one does you don't want to do,” says Shapiro.
“When people talk about their supply chain, you need to find out what are they referring to,” says David Blanchard, Cleveland-based business journalist and author of “Supply Chain Management Best Practices.” “Don't get distracted by groundbreaking work at the biggest companies. Do some benchmarking and find out what works.”
Although giant soup-to-nuts supply chain systems have been devised by major vendors, such as SAP AG and Oracle, “there are not a lot of companies that have the time and resources to set that up and support it. You can't just think of it as a fixed cost and move on. You've got to stick with it every single day throughout the day,” says Blanchard.
On the other hand, you don't need a half-million-dollar supply chain management system to talk to suppliers and customers, he noted. Technology innovation for supply chains often implies a bloodless Rube Goldberg contraption that promised to snuff out human foibles. But many links in the metals supply chain find success with human touch.
Most service center branches have a market radius equal to the distance a truck can drive and return in a day. They know their customers well, because “the service center will keep the customer's purchasing agent out of hot water and keep the production manager out of hot water. Relationships still mean a lot in this business,” says Tenebaum of Chatham Steel.
Software vendors seeking buyers in the metals business say they have learned this lesson. “We had some hard lessons. Now, we try first to discuss their problems instead of our capabilities,” says Verticent's Cranford. “We leave the software until the end of the conversation.”
“Advances in supply chain theory are not sufficiently valuable to overcome price swings” in the metal being bought and sold, says metals industry consultant Sinclair. The most sophisticated sales representative from the most technologically advanced service center can see his or her sales presentation undercut by a competitor with a lower price, he says. In such an environment, nuances in supply chain management technologies pale in significance, he says.
In a typical service center story, Tenenbaum says, customers frequently call service centers to change the terms of scheduled orders, either in regard to quantity, delivery date or metal specifications. Accommodating these changes for the sake of customer service may be profitable for experienced service center managers, but it obliterates opportunities to implement systematic supply chain management practices and forces service centers to carry costly inventories “just in case” instead of turning inventories “just in time.”
Moreover, metal fabricators have little incentive to commit to formal collaboration with any one service center or to share internal information that could reduce their ability to negotiate lower prices and, if the data is leaked, inform competitors.
Despite the complexities and idiosyncrasies of the metals supply chain, one nagging problem draws most of the links together: the need to forecast demand better. Typical metals buyers are small manufacturers that are unable to forecast their demand beyond informal guessing. They can be unwilling to lock themselves into supply contracts much beyond spot buying. Any improvement in demand forecast accuracy by any link in the supply chain can benefit all, if the knowledge is shared.
“The demand-generation side is where the big hurt is,” says Edward Marien, Ph.D., professor emeritus in transportation and logistics management at the University of Wisconsin-Madison's executive education program. “Everybody is working based on demand, but they are all working on their own projections of demand and they are not coordinating. A real step is to look at sharing demand up the supply chain so that all parties are privy to all of that.”
The company culture at many metals companies reflects a fatalistic view of demand forecasts, says consultant Edwards. “It's a crisis every hour. But if you could forecast it, you wouldn't have to worry about it.”
“Open orders are a forecast in a detailed sense,” says Ronald Votto, a sales executive at Quintiq, a Netherlands-based APS software company with North American headquarters in Wayne, Pennsylvania. “If ABC Metals Inc. has an order book to make a certain amount of product X, Y and Z over the next three months, one can assume that the forecast for the next three months is that amount of X, Y and Z. Therefore planning to make X, Y and Z is relatively simple. It's when companies have to look out past the order book, whether it's six months or 18 months for things like material procurement, shift planning, asset utilization, that forecasting becomes a bigger challenge.”
If you can't forecast well, forecast often. “Since we can't forecast very well, we all need to look at the same information so we're not trying to game each other, lie to each other to hedge our purchases,” says consultant Edwards, a fellow alum with Gustafson from ADS Logistics.
Internet-based systems linking supply chain participants can speed demand forecast from end-users to mills. But the forecast inputs don't originate in a computer.
At any level of the supply chain, “there is no substitute for talking to customers,” says Jane Lee, vice president for supply chain solutions at Supply Chain Consultants in Wilmington, Delaware. Up and down the chain, “sales reps call on customers. They have some feel for the way demand is going. We are a software merchant. We sell demand planning software for the supply chain. There are vendors who say you need to take the people element out, but we have found that is not the best way to forecast.”
That's one reason Marmon/Keystone wants its logo on the trucks provided by Ruan Transport even though Ruan provides the logistics and manages the operations. Whenever possible, Ruan retains Marmon's drivers, who are identified with Marmon and are better able to collect informal market forecast intelligence as they deliver, says Joe Ulrich, vice president for sales at Ruan Transport.
“We provide the resources and tools to identify and measure all activities in the delivery process,” Ulrich says. “We provide standardization to the process.” But by retaining a client's truck drivers, “we retain tribal knowledge. My driver is in there talking to customers seeing what the competition is doing. They can feed that back to [Marmon/ Keystone] sales reps.”
The supply chain puzzle has many pieces, some highly technical but many less so. Making all the pieces fit— human and digital—isn't easy. “Rather than throwing up our hands and saying this is too complicated and we can't do anything about it, I would say it's too complicated and we have to do something about it,” says Edwards.