An Electronics Firm Will Save Big Money By Replacing Six People With One And Lose All This Paperwork, Using Enterprise Resource Planning Software. But Not Every Company Has Been So Lucky.
By Michael H. Martin

(FORTUNE Magazine) – It may be the most powerful product endorsement ever: The world's two largest software companies, IBM and Microsoft, run big parts of their business on software that neither one of them makes. Big Blue is well on its way to replacing a hodgepodge of homespun internal software, much of it years in the making, with integrated Enterprise Resource Planning software made by SAP, based in Walldorf, Germany. About 80% of IBM's core business activities may eventually run on ERP systems. Microsoft, meanwhile, has discovered that installing a $25 million system from SAP should produce a common procurement system worldwide that could save $12 million a year in early-payment discounts.

These technology bellwethers, plus other Silicon Valley stars such as Intel, Compaq, and AutoDesk, are relatively recent converts to an ERP religion that has swept through other industries. SAP, the market leader, competes with companies like Oracle, PeopleSoft, J.D. Edwards, and Baan--and business is amazingly good. Some 20,000 companies worldwide paid $10 billion to ERP vendors in 1997, up 40% from 1996. (Investors in the top five ERP vendors averaged a 62% annual return over the past two years.) And that trend is likely to continue. Advanced Manufacturing Research (AMR), a Boston-based firm that specializes in ERP, projects a 36% compound annual growth rate in license revenues through 2001.

The appeal of such integrated information systems to big companies is clear. The sales force enters an order on a computer, and the transaction ripples through the entire company. Inventory lists and parts supplies are updated automatically, worldwide. Production schedules and balance sheets reflect the changes. Best of all, every employee has just the information necessary for the job at hand. Feedback cycles are positive and fast. Salespeople can promise firm delivery dates, and managers can gauge almost immediately the effects of decisions affecting credit terms, discounts, inventory, or supply-chain management.

But there is a problem: making integrated ERP systems perform can be a life-changing experience for a manager because of the high costs and difficulties of installation. That's one reason tech companies were slow to sign on. SAP co-founder Hasso Plattner says that a reputation for complexity trimmed at least three points from his company's remarkable growth rate (50% in compounded annual revenue gains over the past four years). "It's the main thing our competitors play on," he says.

And without doubt, those competitors have plenty of ammunition. Witness the following scene at SAP's recent weeklong industry conference in Orlando. The 12,000 attendees spent their days jawing about technology and their nights carousing at Epcot center and on the Disney-MGM Studios tour. Business reengineering guru Michael Hammer--part management theorist, part carnival barker--gave the keynote. But the most revealing moment may have occurred on a bus full of tipsy sales reps. After brief introductions one rep said to the other, "Hey, I heard you finally got [company X] to sign on the dotted [line]." Second rep: "Yeah. It took a while for them to see the light, though." First rep (laughing): "Now comes the darkness." (Loud laughter throughout bus.)

It's an open secret that the ways to fail at ERP "implementation" (one never simply installs ERP) outnumber the ways to succeed. Horror stories range from technology glitches like Windows desktops that suddenly start speaking in German to heart-stopping charges--and, in the case of a company called Foxmeyer Drug, bankruptcy proceedings. The basic rap is that ERP systems are mind-bogglingly expensive and slow to install. Even a medium-sized installation can soak up tens of millions of dollars and require years of tweaking before benefits appear. Then there's the ERP support industry, an ecosystem of services and consultants whose bills can exceed the price of the software itself by a factor of ten.

While there are as many different methodologies for implementing ERP as there are consultants who will charge 2,000 bucks a day to explain them to you, years of experience yield some basic insights into what works. First rule: Don't expect perfection. A 1996 study by the Standish Group in Dennis, Mass., found that as costs of technology projects go up, the likelihood of failure increases rapidly. "Once you hit $10 million, the chances of a project coming in on time and on budget are statistically zero," says Jim Johnson, Standish's chairman. He estimates that at least 90% of ERP implementations end up late or over budget.

The most important thing to remember is that the real problems with ERP systems like those made by SAP come pretty much from the same place as the benefits--that is, from reengineering a company's core business processes to take advantage of the software. "It's a huge mistake to think about ERP as a technology initiative instead of a business project," says Bruce Richardson, VP for research strategy at AMR. "About 80% of the benefits come from what you change in your business. The software is just an enabler."

The corollary is that any company undertaking an ERP makeover needs commitment to the process from the top levels of management. "You need to get all your senior managers together and talk about the business benefits you're going to be getting," says Tom Davenport, an infotech specialist at the University of Texas at Austin. Davenport, who has studied scores of ERP implementations, explains: "It's an integrated system. It's not going to work if a country manager in France says, 'You didn't involve me.' Everyone's going to want exceptions; everyone's going to say they don't take orders in the same way you do. But that's going to kill the project."

Even with all the top managers pulling in the same direction, it's an excruciating process. "It's like mapping out the entire genetic structure of a human being," says AMR's Richardson. It requires dissecting every link in the operational and decision-making chains and then reconstructing them to take advantage of the new systems.

For Hershey Foods, a $4-billion-a-year company known for its chocolate, reengineering for ERP means extracting and then standardizing into one integrated system the best finance, logistics, and sales practices from multiple divisions. "It's trying to get your business people to concur on how you're going to do this or that process," says Rick Bentz, Hershey's VP for information technology. Bentz oversees the company's three-year, $75 million undertaking. "You think that the problem is solved, and then you get farther down the road and find that it's not. It's two steps forward and one step back."

When it works, it makes for efficiency and better customer service. Because the new computer systems let Hershey track all outstanding accounts and thus measure overall payment risk, credit terms should for the first time be consistent across all Hershey's divisions, large and small. That means supermarkets will no longer have to haggle about credit terms once with the chocolate division for candy bars and again with the pasta division over Ronzoni spaghetti. Hershey also hopes to disseminate "best practices" for key activities like demand forecasting companywide. About 85% of Hershey's core business processes should be running on ERP infrastructure by 2001.

Since many senior managers rebel at the idea that they should be constrained because of an information system, it's vital that companies implementing ERP be able to make decisions quickly and authoritatively. Edith Martin, who oversaw Eastman Kodak's $500 million SAP implementation--or more accurately a reimplementation, since an earlier attempt had failed--says that one key to her success was a council of senior executives to review the project continually and make decisions within days. "We sold them on the economics, we set some dates, and we started working on it," she says. "The idea was to take it out of the dimension of information technology and into strategy. The thinking became that Kodak could not afford to fail at this."

That kind of commitment paid off when some division heads grew impatient. "We had people ready to charge forward before the rest of the company was ready. These were the big guys on the block. They were used to doing what they wanted," she says. But Kodak CEO George Fisher, who sat on the project's executive council, forced them to follow company priorities, not simply what looked best at the division level. "Just because you've got cash in the budget doesn't mean you get to go first," says Martin. "Also, if you've got the CEO in on it, you know they're not going to make an end run."

On the other hand, managers can't do it by themselves. "The easiest mistake to make is underestimating the time and cost of training the end-users," says Sam Cox, chief information officer at Brother Industries USA in Bartlett, Tenn. (the electronics firm mentioned on this article's opening page). Cox has worked on ERP implementations with software from IBM, Mapics, and J.D. Edwards. "Data accuracy is best at its source," he says, citing engineers at an appliance company who felt that data entry was beneath them. "The engineers wanted to work on paper and hire clerks to input the data," he says. "It can be as simple as giving the guys in the warehouse lessons in using a mouse instead of a pencil." Cox currently heads up the Japanese electronics company's $5 million pilot project, which Brother hopes will cut inventory at the $400-million-a-year plant by 18% the first year.

Another problem is that ERP has taken off so fast that finding skilled consultants can be difficult. Companies still complain about visiting techies with only a few months' training who charge $2,500 a day. The obvious solution--training a company's internal staff--is only slightly less risky. Experienced employees are often lured away by competitors.

Under pressure from the vendors to simplify things, many consultants now use business-process templates that prescribe the best practices for particular industries. The big advantage is that many of the bugs that once had to be solved laboriously for each customer are now gone. But templates remain the property of the consulting firm, which means that competitors can get most of the same benefits. In addition, every implementation requires a certain amount of customization, and that's where you need consultants with the most experience.

Companies should ask how many of the consultants working on a given project are certified by the vendor and how experienced are the hands-on worker bees--not just the pitchmen or team leaders. It's worth paying a premium for consultants with knowledge of your industry, and even writing the names of specific managers into the contract.

Sometimes the only way to resolve implementation difficulties may be to scuttle the project. That's what happened at computer direct-seller Dell. CIO Jerry Gregoire recently pulled the plug on an SAP system that was supposed to have connected almost all aspects of the company's operations. Gregoire says that in the fast-paced PC marketplace, it is crucial to be able to make changes in ordering, manufacturing, and other systems quickly, something that is difficult with highly integrated systems.

While emphasizing that Dell was not dissatisfied with the software already installed--in fact, the company will keep some components of ERP--Gregoire says the new plan is to rely on a collection of proprietary and packaged software. The idea is to eventually "run Dell on Dell [hardware]," something that would have been technically impossible with SAP in the near future.

Any company has to view ERP within the context of its overall business. Tom Davenport argues that as ERP becomes ubiquitous, senior execs should consider a new set of questions. "What are the real business processes that bring us our identity and our competitive advantage? And how can we be sure that we really have some unique capability there?" Compaq Computer, for example, maintains its edge by keeping ERP software out of areas like product forecasting.

For most managers faced with the onslaught of ERP, though, the more immediate question is getting from here to there in good order. Hershey's Rick Bentz sums it up well: "I can't wait till all this is finished," he says. "Our biggest risk is taking our eye off our normal business. " Whatever the potential benefits of this new technology, few companies can afford to take that risk for long.