The millennium started poorly for North Face Corp. The company’s logo was ubiquitous on college campuses, yet the manufacturer, distributor, and retailer of outdoor apparel was nearly bankrupt, largely because of problems with its supply chain.
“North Face was losing $100 million on $238 million of sales because it was producing too much inventory and could not match this inventory to orders or even ship it effectively,” says Dan Templin, CFO of Outdoor Coalition, a division of VF Corp., the $5.6 billion apparel giant that bought North Face in May 2000.
“The company had zero visibility into its supply chain,” says Templin. “It didn’t know what its orders were or where its purchases were logistically. Everything had come unglued from an information and service standpoint.” On-time performance to retailers was less than 50 percent, bad for any manufacturer but particularly deadly for North Face because there was enormous demand for its apparel.
Given that even “performance” clothing has the limited shelf life of other fashions, there was little opportunity to recoup one season’s botched opportunities the next year. The company, says Templin, was in crisis mode.
VF saw potential, but it had reservations. “We were looking at a business that could not meet its planned sales, had a broken supply chain, and was losing money fast,” says Templin, who was CFO of VF’s knitwear division at the time. Ultimately, what sold VF on North Face was the latter’s solid brand and VF’s faith in its own operational expertise and the power of business-intelligence (BI) software to fix supply-chain problems.
Templin had seen what BI software could do, having used products from Brio Software to fix supply-chain problems at VF’s knitwear division and throughout the company. Confident that the experience could be repeated, VF bought North Face.
North Face’s supply chain was undermined by its reliance on manual processes, arduous under any circumstances and more so for North Face, which had to cope with a mountain of data. The company created reports on nearly 12,000 stock-keeping units (SKUs) and 2,000 resellers.
Transaction data included sales orders, purchase orders, and inventory, which were entered into spreadsheets from which the reports were created. The effort often took weeks and sometimes months, frequently creating situations in which orders on the front end couldn’t be aligned with all the back-end execution needed to satisfy them. Templin wanted to automate the entire process with a data warehouse and BI tools that could provide visibility into orders and drive smarter purchase decisions.
The Brio software culls data from the warehouse to produce more than 50 electronic reports that compare sales-order information with inventory data, determining an accurate delivery date that meets or precedes a customer’s requested ship date. Reports cover historical sales information and open sales orders by customer; the styles, colors, and sizes that are best-sellers and those that may require discontinuation; and more-traditional metrics like business-unit profit-and-loss statements and budgets versus actual spending. The reports are pushed out to North Face decision-makers via E-mail.
By monitoring those 12,000 SKUs along the entire supply chain, North Face has reduced excess inventory by 65 percent. The company now fails to meet just 2.5 percent of its orders, compared with 10 percent before VF and BI came to the rescue.
And while most retailers would view store closures as a sign of failure, in this case, the company is cheering. With so little excess inventory to unload, the company has closed 13 of its 16 outlet stores. Unloading so much of its own stuff at a discount was beginning to erode the brand, says Templin. That may be bad news for college kids on a budget, but it’s good for North Face, which is now seeing higher margins, regularly hitting the 14 percent profit margin that VF demands.
All of this may come as a surprise to CFOs and other longtime users of BI products, who have tended to regard the software as great for analyzing internal financials but not particularly well suited to supply-chain issues.
But a study last year by research firm International Data Corp. found that many companies applying BI software to supply-chain problems are realizing a hefty return. IDC studied analytics projects undertaken to address three core business processes — customer relationship, financial/business performance, and operations/production, which includes supply-chain management — and found that, when projected over a five-year period, supply-chain projects actually showed the highest return. “The median return for operations/production analytics was 277 percent, far more than the median return of 139 percent for financial/business performance management and 55 percent from CRM applications,” says Henry Morris, IDC’s group vice president of applications and information access.
“This is a very fruitful area,” contends Morris. “Supply-chain analytics can boost the bottom line because they produce greater efficiency, less scrap, better quality, and lower production costs, and can improve the top line through greater customer satisfaction. This is basic business made better.”
Keith Pratt, CFO of Advanced Fibre Communications, would agree. In 2001, he put pen to paper and wrote off $30 million in excess inventory and overbooked purchase commitments to a variety of its suppliers. For a company like AFC, a designer and manufacturer of broadband access solutions with 2002 revenues of $344 million, $30 million was a tough nut to swallow, especially for something as seemingly controllable as inventory.
AFC certainly wasn’t alone in suffering the boom-bust supply-demand cycle that suddenly manifested itself in 2001. Other companies servicing the telecommunications industry, notably Cisco Systems, also made substantial purchase commitments to suppliers in 2000 based on strong customer demand and constrictions in their suppliers’ capacity to meet the demand.
When demand fell and sales plummeted, many firms were left holding the bag, which in AFC’s case amounted to a hefty $18 million in excess inventory and another $12 million in commitments to suppliers to buy more components. “We were unable to respond to demand changes in time to prevent being overloaded with inventory,” says Pratt. “We needed a way to respond quickly to changes in demand and have it percolate through our supply chain.”
As part of the solution, AFC turned to BI software company Valdero, hoping to open a window into its suppliers’ supply chains. It now has visibility into the purchase commitments made by its four top-tier suppliers with their respective second-tier suppliers, dollar commitments that AFC ultimately is liable for. If demand falls as it did in 2001, AFC is ready with an updated forecast and assurances that tier-one suppliers are taking appropriate actions.
Thanks to the BI tool, AFC’s materials planning more accurately reflects demand. “We’ve improved inventory turns from three a year to seven a year, meaning what used to be on our books for four months now is on our books for less than two months,” the CFO says.
A Fine Mesh
But there is a spate of activity as traditional BI vendors begin to offer features that analyze the supply chain, traditional supply-chain integration vendors begin to add BI features, and ERP vendors look to do some of both. Each has its own take on how to pull and push data from front-end demand forecasting to back-end execution — not to mention all the back-end integration required, plus the connections to financial metrics so that glitches in the supply chain can be cast in a context users of the software can appreciate.
There is a compelling financial reason for using BI tools to gain visibility into the supply chain: “You want to mesh your anticipated demand with your ability to meet that demand accurately and quickly, without getting stuck with too much inventory or not enough,” says David Folger, vice president in the enterprise analytics strategies service of Stamford, Connecticut-based research firm Meta Group.
And, Folger adds, users of BI tools that tackle the supply chain also want to be able to “glean exogenous events that may prevent a supplier from meeting your needs and pull all this data together across the enterprise, so that it’s updated and monitored to alert you when things are out of synch.”
Those are lofty goals, most of which have yet to be met. Companies deploying BI software that addresses the supply chain are often constrained by their own legacy systems, disparate systems used by suppliers, and the fact that vendors possess different areas of expertise.
Data integrity (customers and salespeople often offer bloated projections) and supplier resistance (“Whaddya mean you want to open a window into my supply-chain system?”) also pose challenges. For now, companies must settle for tactical solutions provided by the disparate vendors, pieced together to offer a semblance of integration. “The Holy Grail is real-time, accurate visibility and predictability, where a demand forecast feeds into orders for supplies, and supplies are monitored to ensure they meet this demand,” says Amir Hartman, managing director and co-founder of consulting firm Mainstay Partners and a senior fellow at Harvard Business School. “But we’re not there yet.”
Even North Face, with all its success, does not have a true end-to-end integrated supply chain. For example, unlike AFC, North Face lacks an automated window into its suppliers’ supply chains. “We haven’t decided how we would go about that yet,” says Templin, “or which vendor would provide it. But integration across the supply chain is a strategic goal and part of our philosophy.”
AFC says that it, too, is missing certain pieces. “What we’re trying to do now is get better forecasting and planning tools that we can hook up to the back-end systems,” says Jeff Rosen, who heads AFC’s IT organization as vice president of operations. “We want to make sure suppliers at the back end meet the forecasted demand accurately and that inventory levels are appropriate. That’s the burning issue.”
Progress Rail Services, on the other hand, tackled the front end first. The Albertville, Alabama-based company, which makes signal-crossing and track switches, repairs locomotives, and offers maintenance services, faced an internal train wreck: as a result of 20 to 25 acquisitions in a three-to-five-year span, it found itself using multiple financial systems, which made analysis a very long haul. “I desperately needed transparency across these various systems to manage our assets and supply-chain information,” says CFO David Klementz, “as well as confidence in the integrity of data, given all these systems and spreadsheets.”
He turned to BI vendor SAS Institute. “I’m able now to look at performance drivers by business unit or division, to make assumptions about forecasting and managing costs, and to manage inventory levels,” says Klementz. Financial and other data from those multiple systems are consolidated in a data warehouse, on top of which sits the BI solution. “I have my forecast and actual financials in one place, which ultimately tie into my scorecard of performance indicators and provide a directional view of where we’re headed,” he adds.
But more important, Klementz now has a grip on the company’s inventory levels relative to the forecast. “If the economy teeters or there are other changes in the business environment, I can adjust the inventory strategies,” he explains. “Instead of making intuitive business decisions with my inventory or other financial indicators, I make fact-based decisions. If I have what now appears to be too much inventory, I can increase the sales focus or strategy in that area to push it harder, knowing that otherwise, it may carry inventory longer than desired.”
The BI tool was implemented last fall, and additional phase-ins remain, so Klementz does not yet have actual metrics indicating how well it has performed. But he’s confident it will significantly improve business performance, and, in particular, inventory turns.
He hopes someday to integrate BI across the enterprise to create competition among the company’s suppliers. “I need more visibility to tell what we’re buying from whom, in what quantity, and at what price,” he says. “Right now, I can’t easily tell if I’m buying similar functioning items with slightly different specifications from 10 different suppliers, and if I bought it from just 1 whether or not I could save money.”
All these companies would love a one-size-fits-all BI solution that integrates all the links in their supply chains with other IT systems and finance, a process that in large part consists of watching the universe of software makers to see which offers what when. “I would think the supply-chain vendors [i2, Manugistics, and others], given their domain knowledge, will define the standards for full integration,” says Meta Group’s Folger. “Then again, the ERP vendors are so much bigger and so financially healthy, and the BI vendors are the early movers with analytics capabilities. I imagine they will all continue to play a role.”
IDC’s Morris also finds it difficult to pick a winner. “The ERP vendors as presently constituted understand business processes and offer BI tools, but they don’t address all the links in the supply chain,” he says. “The supply-chain vendors are behind the curve as far as analytics, as they are weak in the areas of forecasting and trend analysis. Meanwhile, the BI vendors, which have the lead as far as turning data into informative reports, historically have offered technology that is not well integrated to business processes. What is needed are solutions that integrate ERP, supply-chain management, and business intelligence that is event- and business-process-aware.”
Folger warns against waiting for a technology cure-all. “Many companies want a system where the forecast feeds into the execution side with little or no human involvement,” he explains. “This way, when the forecast suffers, supplies and inventory automatically reduce, and when they rise, supplies and inventory increase. But if you’re the guy running procurement and things are booming, and some BI tool indicates a bust is around the next corner and you should cut procurement dramatically, you’d be an idiot to accept it at face value. Imagine going to the CEO and saying we’re cutting production even though we’re making more money than ever.”
The moral? “The human element will always be necessary,” says Folger. “You can’t completely automate everything.” But companies that have brought BI to bear on supply-chain challenges say that even less-than-perfect approaches have paid off, and t hey expect things to only get better.
Sidebar: Where the Rubber Meets the Code
Goodyear Tire & Rubber also wanted better forecasts so that its supply chain was in synch with sales expectations in North America and Europe. A data warehousing tool from Teradata and a BI tool from Cognos measure forecasts against actual sales to pinpoint the gaps. “We wanted to know where our strengths and weaknesses were vis-à-vis the forecast,” says Eric Berg, CIO of the Akron, Ohio-based company. “We also wanted to compare the forecast to our fill-rate measures — whether or not we were filling orders on time. If we were not hitting the 100 percent fill rate, then we’d swing back to the forecast to see if it was accurate. If it wasn’t, we would find out why — for example, a customer could have ordered a different quantity or mix of products than originally anticipated — and then design corrective measures.”
The overarching goal, he says, has been to gain greater visibility into the supply chain to make better business decisions, anticipating and meeting customer demand at the lowest possible cost.
Key performance indicators pertaining to the supply chain are linked to Goodyear’s financial-performance indicators — for example, a fill-rate issue with a customer that may have an impact on revenues. Says Rich Kramer, Goodyear’s vice president of finance for its North American tire business, “By linking daily financial performance to the related key performance indicators across the business and in the supply chain, we can take corrective action immediately, versus waiting for month-end results.”