July 1, 2005


For many metals companies, there are few greater headaches than choosing and implementing enterprise resource planning software. What happens between the promise of ERP and the wide-awake reality?

Eugene H. McNichols, chief executive officer of McNichols Company, the metals service center operator based in Tampa, summarizes succinctly his company's first experience with enterprise resource planning (ERP) software from JD Edwards:

“After nine days of heck, we pulled the plug,” McNichols says. “We had a crisis room that we set up before we went live, and it was frightening to go into that room. Nothing was working. It was just one mistake on top of another, one person in trouble after another.

“I called a meeting. We held a vote. I asked who wants to stay with this stuff. One guy did. We fired him.”

Ask a metals industry CEO about their experience with ERP and other complicated business management software and chances are excellent that you'll hear a story filled with pain, huge investments of time, confusion, critical business functions that stop working and bafflement. Also, crane loads of money. In McNichols' case, for example, the cost of the nightmare, generic JD Edwards ERP system was “millions and millions, and we are currently running only 10% to 15% of what we bought,” he says of his 17-location service center company specializing in perforated and expanded metals, wire cloth and all types of grating and flooring.

Says Bill Jones, president and chief executive officer of O'Neal Steel Inc., the Birmingham, Alabama operator of a network of service centers and a company that has endured its share of ERP angst, “All software upgrades will cost you more than you thought. They will take longer to implement than you thought, and you will probably get fewer benefits than you thought.”

Why has the metals industry experience with ERP and management software often been so problematic? What happens between identification of the need for a technical solution to business management and information issues and the actual implementation leading to such dim results? Several answers emerge from conversations with metals companies and software providers.

FIRST, and most important, most generic ERP packages (such as SAP, JD Edwards, Oracle, Baan and others) are designed with an inventory management system based on discrete, fungible units, such as a standard-sized screw, a specific part, or the model number of a finished product. Consequently, most generic ERP systems find it very difficult to keep track of inventory or compute costs for products, like metals, that are transformed through one or more fabrication processes from basic forms, such as a coil, into a practically infinite variety of shapes, sizes and parts.

“The basic inventory management system is wrong, because steel doesn't replenish the way that other parts do,” says Heber MacWilliams, chief information officer of Olympic Steel Inc. of Bedford Heights, Ohio, and a former professor of management information systems at Case Western Reserve University. “Every piece of steel that you sell can be different from every other piece, in multiple ways.”

SECOND, a piece of metal in a service center's inventory is identified by a dozen or more attributes – coating, thickness, length, chemical properties, metallurgical characteristics, end-user requirements. Generic ERP systems aren't designed to track attributes. “You can't describe in one part number all of those attributes and properties, and some well-known companies in the industry have tragic stories to tell about when they tried to do so with their generic ERP systems,” MacWilliams says.

This was the case with McNichols, which found that the software from JD Edwards, now part of Oracle Corp., was developed to serve the oil and gas industry, and thus didn't fit the specific parameters of the metals business. “No one ERP vendor, among the big companies, develops anything other than ‘one size fits all' software,” says Herbert T. Goetschius, president and chief operating officer of McNichols.

THIRD, even industry-specific ERP systems (such as STEELMAN, OpenTrac, Enterprise 21, ERP Plus, or AXIOM) may not yet provide sufficient support for all of the most complex processes that larger service centers increasingly perform. A service center may convert a coil to a sheet, or a plate into many kinds of parts. One plate may be used to make parts for multiple customers. Customers may require kits of parts, packed in the order they will be needed on the shop floor. Parts can be fabricated at one location, or the metal may be sent by a service center to another company for specialized processing. Very often, tracking such activities requires custom programming.

The assumption of generic ERP is that the work center performs a single assembly or fabrication process, but service centers also have transformation processes,” MacWilliams says.

“The generic ERP system, because of its very nature, is intended to make one thing from many things,” says Mark Alfred, a support consultant for AXIS Computer Systems Inc., provider of AXIOM, a well-known metals-specific ERP application. “But when we are dealing with metals, we are talking about the opposite, one thing converting to many things. That's a big paradigm shift for an ERP system.”

JD Edwards, for instance, has no capacity to track remnants. “We cut metal to size for our customers, and we want to inventory our significant remnants,” says Goetschius, “but JD Edwards didn't have a way to do that.”

FOURTH, any new system means changing long-ingrained work habits. What appears on screens doesn't look the same as it did with the company's legacy systems. Commands change. Information pathways are altered.

“We made a lot of compromises in sales software design to minimize the changes in the JD Edwards software which then would become a maintenance problem when the software is upgraded to the next version,” Goetschius says. “That met a lot of technical needs, but it was not as customer- or user-friendly. The sales people weren't happy. It was too clunky for them.” Says Peter Weymouth, AXIS production systems product manager, “People want exactly what they want, the way they want it. They cling to those features and functions that they developed in the home-grown system. That's why it is so difficult to find an off-the-shelf package that fits.”

FIFTH, and for experienced business people this almost goes without saying, engaging the services of apparently competent consultants is no guarantee that the project will go well. McNichols, for example, worked with a well-known consultant that advised the full implementation of the JD Edwards system, all modules, simultaneously. “It was such a large number of programs at one time that it overtaxed our hardware,” Goetschius says. “Elements of the program when tested in isolation worked fine, but when tested as part of a full ERP system, with other modules, it taxed the system to failure.” In the end, McNichols retained only the system's accounting capability.

O'Neal at one point sued Andersen Consulting for what it said was a failed engagement of business management software.

“We need to make sure that we understand, internally, everything so that we can continue to improve on a system once we have it,” says Jones of O'Neal Steel. “If you relegate or delegate to consultants, when they leave, your knowledge base leaves with them.”


So, what's the answer? To begin with, understand that each type of ERP solution – generic or industry-specific – comes with its own set of benefits and limitations. “Generic ERP packages come with functionality that I would love to have in industry-specific packages,” says MacWilliams, including sophisticated manufacturing and fabrication support, routing capabilities and work center controls. Industry-specific software, on the other hand, is built to work with metals products – obviously of critical importance – but can be weaker in tracking the processes that many service centers use today.

So, for some companies, the solution is to use a sturdy, generic ERP system as the framework for creation of their own business management and information software. The downside is that upgrades provided by the original ERP vendor will, at some point, no longer apply to the company's customized software. The upside is that by building its own systems, the company can include features that ERP companies, with their more generalized view of business needs, might not even know would be useful. For other companies, often smaller metals businesses that run fewer fabrication processes, industry-specific ERP software may resolve most management and information issues.

“Either approach has its own set of problems that will have to be addressed,” MacWilliams says. “The important thing is to understand that and to be ready to respond.”

At Yarde Metals, for example, Chief Technical Officer David W. Pippenger heads a 10-member Management Information Services team that has developed its original business management software, MetalTraq, into a sophisticated, integrated, Linux-based business management system that fits the $300 million company like a digital glove.

“There have been thousands of changes made over the last 15 years,” Pippenger says. “It's a continuous process of improvement. People have little tweaks to the system, sometimes big ideas, and we put it in place.”

For example, Yarde, of Southington, Connecticut, built its own bar code functionality, and is moving to a “locationless” inventory, where products are scanned into a specific area of the warehouse, eliminating the need to repack material. Yarde's MIS department built a telephone system, completely Internet-enabled, for the company's new headquarters office. The company is completely Web-enabled, with home-grown solutions.

Although the MIS department is comparatively large, the cost of some parts of the system are small. When the company requires new computer terminals, it buys ancient, used Pentium II-based personal computers by the pallet-load for $30 each.



Here are a few important points to keep in mind as your company considers business management/information options:

  • ERP-style software can be extraordinarily helpful.
  • Such software comes in generic and industry-specific varieties, each with its own set of difficult issues for metals companies.
  • It is critically important to do your own homework on likely benefits and problems with the anticipated system. Use consultants, but don't rely solely on their advice
  • Clearly understand where your business fits in the spectrum of size, complexity and goals.
  • Always purchase access to the source code of any software package.
  • Always have available your own IT capability to produce customized workarounds and company-specific functionality.

“The beauty is the cost savings for us,” Pippenger says. “We are completely self-provisioned and self sufficient.”

“Roll your own” IT solutions provide another benefit, says Pippenger. They provide authentic strategic advantages that can't easily be matched by companies that use off-the-shelf software.

“No one wants to hear this,” Pippenger says, “but to make ERP systems work, you really have to look at IT not as a digital janitor service, but as an engineering discipline. People always want to go with Microsoft and have it all in a shrink-wrapped box, everything interchangeable and heuristic. You can run a pretty good business within those parameters, but you give up the possibility of doing important things that advance the business in return for the safety of having a vendor to call.”

Absolutely true, says David Sinclair, vice president of information technology for Niagara Corporation, parent of cold finished steel bar fabricator Niagara-LaSalle Corporation of New York City.

“What most people don't understand is that ERP standardizes you and your processes,” he says. “It gives you a single version of the truth, but it doesn't give you a competitive advantage, because everyone else is also implementing the same management software.”


Like Yarde, Niagara-LaSalle runs another older ERP system, Symix, which is no longer published, but after 10 years of continual modification by the company's own IT department, it is 85% customized. “Basic Symix was unable to cope with the growing size and complexity of the business as additional companies were acquired. “The only thing for IT to do in that kind of situation is to acknowledge that the system doesn't work,” Sinclair says. “Then you look for the pain points where you can do a lot of good in a short period of time.”

Symix, says Sinclair, “had no notion of a heat, no notion of all those attributes of steel.” Part of customizing the off-the-shelf system was to craft new reports to provide those attributes and other critical information to the sales force. That quick improvement also bought the IT staff time to work on other underlying issues.


“You ask yourself where your functionality is weak, where the integration is not as strong as it could be,” Sinclair says. For example, Niagara-Lasalle's evolving system lacked adequate transportation functions. It took three to five days to build a truckload for shipping with inventory validated, bundles split and everything ready to go.

“To build truckloads you have to know what's available to ship, where it can ship, and you have to build loads efficiently,” Sinclair says. “ERP systems out of the box will not do that. They aren't going to dispatch loads through a common carrier.” The system Sinclair's team built reduced load-building to a one-day process. Using its home-grown transportation system, Niagara-LaSalle has maintained costs, from a base rate perspective, for the last four years, “and there is no other part of the business where you can say that.”

Order entry time has been reduced by 30%. Niagara-LaSalle is rated No. 1 or No. 2 in terms of delivery performance by all customers who report to the company. In part because of IT improvements, operating costs have dropped 20% partially because of IT.

O'Neal Steel never sought an ERP system. Rather, it wanted the benefits of one. “We were looking for better information, better communication and better productivity for our business,” Jones says. The company evaluated industry-specific software packages and concluded none fit its needs. “Our company is a hybrid of distribution, processing and manufacturing in 40 different operations,” says Jones. “No one industry-specific software addresses all the areas that we need.

“The difficulty with ERP comes in the modification and the interfacing necessary for it to fit your business. But we still think that's the best option that we have.”

Although O'Neal underestimated the cost of a management system, the complexity of the task and “resistance to change” from the staff, the company has learned other valuable lessons about itself.

“We also underestimated our ability to change,” says Jones. “Once you get beyond the first resistance, our people have done a good job with new procedures and new ways of thinking about our business. The system doesn't make you do that. It is simply a tool that allows that. You can't live without it.”



ERP software companies maintain that today's industry-specific ERP systems can deal with the metals-industry market environment.

But the key for any potential ERP customer is to thoroughly check out the experience of other companies with the contemplated package.

“You have to rely very heavily on ferreting out what users at other companies tell you about the system,” says Herbert T. Goetschius, president and chief operating officer of McNichols Company. “Don't just rely on the salespeople from the software companies, or even the consultants, who all have biases in favor of certain packages of software. You have to go out, talk with people, find out what works and what doesn't.”

That said, a number of companies that specialize in the metals industry offer solutions that address many of the specific needs of a service center or mill.

AXIS Computer Systems, publisher of the AXIOM ERP system for the metals industry, offers ERP software that defines metal inventory in terms of attributes—shape, size, color and not by part numbers. AXIOM reports on the chemistry of every item in a shipment or an inventory, and it gives the complete transformational genealogy of a heat and all the parts made from it. A planned improvement is to replace a rigid system of preplanned processing routines with dynamic routing, a flexible process that adjusts to fit the particular part and customer requirement, order by order.

“You can print out any physical properties, grain size, yield, tensile strength, elongation information, with the ability to download information right from the mill,” says Bill Swartz, an AXIS senior consultant. AXIS Director of Sales Ronald J. Greco says, in fact, that the typical company finds that 85% of what they need is available from AXIOM, “right out of the box.”

So if today's ERP systems can be so flexible, why are they so difficult to implement? AXIS after all, says it takes six to nine months, at best, to implement AXIOM, and a major job can easily last 18 months.

A major issue, says Greco, is culture shock – new ways of using computers that impact every aspect of the business.

“People don't understand the scope of an implementation,” he says. “It hits all four corners of an organization, and it touches every part of the business. It happens while your people are running the business, and now they have to implement software at the same time. Most metals companies don't have people sitting around waiting to put a computer system in.”

Clinging to old ways was not the issue for Germany's Salzgitter AG, a maker, fabricator and distributor of flat-rolled steel and steel tubes. Salzgitter's IT organization, known as GESIS, elected to mold an SAP system for the Salzgitter Trade Company to create a unified product database, order-entry and inventory system for the company's 25 European plants, permitting Salzgitter to draw on all of its resources to respond to customer requirements anywhere on the continent.

“We have one SAP system for the steel trade which has its own additional developed modules, and another system that controls production,” says Bruno Ritter, a GESIS senior consultant, “and with the SAP system, we have nearly finished creating a cross-supply chain between the Salzgitter Trade Company and the Salzgitter Steel Making Company, and between the sales organization and our customers and suppliers.”

But it wasn't easy. Salzgitter has had a 20-year relationship with SAP, and a major issue for the new trade system was the 25,000 product variables that the system had to recognize.

“We had to develop unit conversions as one variant,” Ritter says. “It wasn't available, and we wanted it. So we worked with SAP. We needed pipe to be recognized in meters, kilograms and pieces, all with different dimensions, all with formulas to calculate cost and profit. For beams, we had to have dimensions, length and quality information. We needed product certification with every batch of material.”

The first implementations began in late 2003, and the first Salzgitter facility to use the system went live with it in February 2004. “Yes, we had some bad days,” Ritter recalls. “System performance was too slow. We had to upgrade some hardware and adopt new routines. But now we have a system that responds in two to three seconds. Now, for our 25 plants we have open storage for the seller. Any plant can tap the inventory of any other plant and for the customer, it doesn't matter where the material is.”

The effort was beneficial for SAP, too. New tricks learned by SAP from the Salzgitter project have been incorporated into the software company's mill products application, developed specifically for metals companies.