“I thought we weren’t going to talk about i2,” growls Roland Wolfram, Nike’s vice president of global operations and technology, his eyes flashing at his PR manager with ill-concealed ire.
Wolfram, who was promoted in April to vice president and general manager of the Asia-Pacific division, is all Nike. His complexion is ruddy, his lips cracked from working out or working hard, or both. He’s casually dressed, but with a typical Nike sharpness to his turtleneck and slacks, a sharpness reflected also in his urgent, aggressive defense of his company—a Nike pride that would seem arrogant were not the company so dominant in its industry.
Wolfram calls the i2 problem—a software glitch that cost Nike more than $100 million in lost sales, depressed its stock price by 20 percent, triggered a flurry of class-action lawsuits, and caused its chairman, president and CEO, Phil Knight, to lament famously, “This is what you get for $400 million, huh?”—a “speed bump.” Some speed bump. In the athletic footwear business, only Nike, with a 32 percent worldwide market share (almost double Adidas, its nearest rival) and a $20 billion market cap that’s more than the rest of the manufacturers and retailers in the industry combined, could afford to talk about $100 million like that.
It drives Wolfram crazy that while the rest of the world knows his company for its swooshbuckling marketing and its association with the world’s most famous athletes, the IT world thinks of Nike as the company that screwed up its supply chain—specifically, the i2 demand-planning engine that, in 2000, spat out orders for thousands more Air Garnett sneakers than the market had appetite for and called for thousands fewer Air Jordans than were needed.
“For the people who follow this sort of thing, we became a poster child [for failed implementations],” Wolfram says.
But there was a lesson too for people who do, in fact, follow “this sort of thing,” specifically CIOs. The lesson of Nike’s failure and subsequent rebound lies in the fact that it had a business plan that was widely understood and accepted at every level of the company. Given that, and the resiliency it afforded the company, in the end the i2 failure turned out to be, indeed, just a “speed bump.”
The i2 Failure: Tactical or Strategic?
Nike’s June 2000 problems with its i2 system reflect the double whammy typical of high-profile enterprise computing failures. First, there’s a software problem closely tied to a core business process—in this case, factory orders. Then the glitch sends a ripple through product delivery that grows into a wave crashing on the balance sheet. The wave is big enough that the company must reveal the losses at a quarterly conference call with analysts or risk the wrath of the Securities and Exchange Commission, shareholders or both. And that’s when it hits the pages of The Wall Street Journal, inspiring articles and white papers on the general subject of IT’s hubris, limitations, value and cost.
The idea that something so mundane as a computer glitch could affect the performance of a huge company is still so novel that it makes headlines. But what doesn’t usually enter the analysis is whether the problem was tactical (and fixable) or strategic (meaning the company should never have bought the software in the first place and most likely won’t ever get any value from it). The latter is a goof worthy of a poster; the former is a speed bump.
Nike claims that the problems with its i2 demand-planning software were tactical and therefore fixable. It was too slow, didn’t integrate well, had some bugs, and Nike’s planners were inadequately trained in how to use the system before it went live. Nike says all these problems were fixed by fall 2000. And the company asserts that its business wasn’t affected after that quarter. Indeed, at press time, Nike had just announced that its third-quarter 2003 profit margins were its highest ever.
If there was a strategic failure in Nike’s supply chain project, it was that Nike had bought in to software designed to crystal ball demand. Throwing a bunch of historical sales numbers into a program and waiting for a magic number to emerge from the algorithm—the basic concept behind demand-planning software—doesn’t work well anywhere, and in this case didn’t even support Nike’s business model. Nike depends upon tightly controlling the athletic footwear supply chain and getting retailers to commit to orders far in advance. There’s not much room for a crystal ball in that scenario.
Indeed, Nike confirms that it stopped using i2’s demand planner for its short- and medium-range sneaker planning (it’s still used for Nike’s small but growing apparel business) in the spring of 2001, moving those functions into its SAP ERP system, which is grounded more in orders and invoices than in predictive algorithms. “This allows us to simplify some of our integration requirements,” says Nike CIO Gordon Steele.
Wolfram says Nike’s demand-planning strategy was and continues to be a mixture of art and technology. Nike sells too many products (120,000) in too many cycles (four per year) to do things by intuition alone. “We’ve tuned our system so we do our runs against [historical models], and then people look at it to make sure it makes sense,” he says. The computer models are trusted more when the product is a reliable seller (that is, just about anything with Michael Jordan’s name on it) and the planners’ intuition plays a bigger role in new or more volatile products. In this case, says Wolfram, talking with retailers does more good than consulting the system.
“There’s been a change in the technology for demand planning,” says AMR Research Vice President Bill Swanton, who declined to address the Nike case specifically. “In the late ’90s, companies said all we need is the data and we can plan everything perfectly. Today, companies are trying to do consensus planning rather than demand planning.” That means moving away from the crystal ball and toward sharing information up and down the supply chain with customers, retailers, distributors and manufacturers. “If you can share information faster and more accurately among a lot of people, you will see trends a lot sooner, and that’s where the true value of supply chain projects are,” Swanton says.
If You Have a Game Plan, You Can Snag the Rebound
Another thing that makes Wolfram angry (his already ruddy complexion going completely red) is the widespread assumption that Nike was betting on algorithms and changed course when that didn’t work out. Wolfram says that, on the contrary, i2’s demand-planning software was never intended to be the hero of Nike’s supply chain project—one of the most ambitious ever attempted by a company its size. It was (and still is), he claims, part of a wider strategy to integrate ERP, supply chain planning and CRM software onto a single platform shared by Nike operations in North America, as well as Europe, the Middle East and Africa (EMEA). “Frankly,” he asserts, “we pretty much stayed the course.”
Nike made a bold early bet on the risky and difficult strategy of creating a single, giant, integrated database within its SAP ERP system for every employee in North America and EMEA. (Nike’s Asia-Pacific division will be on a separate instance of the software.) This meant getting everyone to agree on business practices and common data definitions before the software went in—a rarity in ERP project management.
The difficulty of integrating information across a distributed company has brought down many ERP projects, such as drugstore chain FoxMeyer’s SAP ERP system in the late ’90s and Tri-Valley Growers’ 1997 choice of Oracle’s ill-fated ERP package for the consumer packaged-good industry. Neither company ever got its systems working properly and that contributed to both eventually shutting their doors. Other companies gave up on the vision of total information integration and installed many different versions of their ERP systems—as many as 400 different versions, or instances, of a single vendor’s ERP system at some really large companies, according to AMR.
But Nike claims it has never wavered from its single-instance strategy, even when problems with the first piece of that strategy, the i2 system, hit the news on Feb. 26, 2001. The same project leaders who were in place at the time of the i2 problems (CIO Steele and the business lead, Shelley Dewey, Nike’s vice president of supply chain) are still running the project today. The reason Steele and Dewey survived was because when their system failed, they had a lifeline to hang onto: a clear business case for the overall supply chain project. If achieved, they claim it will save the company a lot more than Knight’s $400 million and the $100 million in wayward sneakers.
Nike’s supply chain project is supposed to drive the manufacturing cycle for a sneaker down from nine months to six. Cutting out that three months would match Nike’s manufacturing cycle to its retailers’ ordering schedule—they order 90 percent of their sneakers six months in advance of delivery. This means Nike could begin manufacturing its sneakers to order rather than three months in advance and then hoping they can sell them. Converting the supply chain from make-to-sell to make-to-order is the dream of any company desirous of gaining competitive advantage through its supply chain. Dell has done it, famously, with PCs; Nike wants to do it just as famously with sneakers.
Nike hasn’t gotten there yet. And its business case relies on a nearly 30-year-old model that some analysts and retailers grumble is out of touch with the reality of today’s market. But it’s a business case Nike’s leaders believe in. This is how CIOs keep their jobs when a project goes off track and it’s how they keep getting funding to keep it going.
Like many truths, this one is simple yet profound: Projects that survive breakdowns do so because everyone in the business, not just IT, understands what the system is supposed to do for the company—and sees value in it. Indeed, after his infamous conference call outburst in 2001, Knight added that, “I think it will, in the long run, be a competitive advantage.”
“We wish to God Phil [Knight] hadn’t said what he said,” says Steele with a laugh. “But his belief in this project has never wavered. [When the i2 problems emerged], we sat down and talked about what the issues were and he said, OK, I understand, carry on.” (Knight declined to be interviewed by CIO.)
How Nike Built a Robust Business Case
Knight, not normally known for self-control, has shown extraordinary patience with Nike’s supply chain project. And he’s needed it. “Once we got into this, we quickly realized that what we originally thought was going to be a two-to-three-year effort would be more like five to seven,” says Wolfram.
It’s been six years now and counting, with the final stage of the project due to be finished sometime in 2006 at a total cost that has gone from a projected $400 million to $500 million, according to Wolfram.
The theme of Nike’s sneaker supply chain is centralization. All product design, factory contracting and delivery is planned and coordinated from Beaverton, Ore. The supply chain is built around a six-month order cycle, called the “Futures” program, that was developed in 1975 in response to the then-chaotic market for running shoes. In those days, the Far East sneaker supply chain was in its infancy, deliveries were spotty, inflation was high, and runners bought whatever shoes they could find regardless of brand. Nike won that market by guaranteeing delivery and an inflation-proof discount in return for getting its orders six months in advance. Retailers went along happily because runners didn’t much care about style or looks—they wanted technically advanced shoes that fit and were in steady supply. Retailers knew their Nike shoes would sell no matter how far in advance they ordered them.
But as Nike became increasingly global, its supply chain began to fragment. By 1998, Nike had 27 order management systems around the globe, all highly customized and poorly linked to Beaverton. To gain control over its nine-month manufacturing cycle, Nike decided that it needed systems as centralized as its planning processes. ERP software, specifically SAP’s R/3 software, would be the bedrock of Nike’s strategy, with i2 supply, demand and collaboration planner software applications and Siebel’s CRM software also knitted into the overall system using middleware from STC (now SeeBeyond).
Nike’s patience was a virtue here too. It skipped AFS (Apparel and Footwear Solution), the initial version of SAP’s R/3 software developed specifically for the apparel and footwear industry. Archrival Reebok, which partnered with VF (makers of Wrangler Jeans and Vanity Fair bras, among other things) on the beta effort to develop AFS beginning in 1996, struggled for years to implement the buggy, unstable AFS software. (Reebok declined to be interviewed for this story.) And although Nike purchased AFS in 1998, it didn’t attempt to install it until SAP began working on the second, more stable version of the software. “Most of the early adopters were busy installing AFS in 1999,” says Steele with a satisfied smile. “That’s when we began spending a lot of time with SAP, sending our people over to Germany to tell them what we’d like to see in the second version.”
Why I2 Went Wrong
Unfortunately, Nike didn’t apply that same patience to the implementation of the first part of its supply chain strategy: i2’s demand and supply planner software applications. Rather than wait to deploy i2 as part of its SAP ERP project, Nike decided to install i2 beginning in 1999, while it was still using its legacy systems.
According to court documents filed by Nike and i2 shareholders in class-action suits, little went right before June 2000. i2’s predictive demand application and its supply chain planner (which maps out the manufacturing of specific products) used different business rules and stored data in different formats, making it difficult to integrate the two applications. The i2 software needed to be so heavily customized to operate with Nike’s legacy systems that it took as much as a minute for a single entry to be recorded by the software. And, overwhelmed by the tens of millions of product numbers Nike used, the system frequently crashed.
But these problems would have remained only glitches had they not spilled over into factory orders. The system ignored some orders and duplicated others. The demand planner also deleted ordering data six to eight weeks after it was entered, making it impossible for planners to recall what they had asked each factory to produce. Soon, way too many orders for Air Garnetts were going over the wires to Asian factories while calls for Air Jordans were lost or deleted.
When the problems were discovered, Nike had to develop workarounds. Data from i2’s demand predictor had to be downloaded and manually reloaded into the supply chain planner by occupying programmers, quality assurance personnel and businesspeople whenever the applications were required to share data—which was as often as weekly. Consultants were brought in to build databases to bypass portions of the i2 applications, and custom bridges were constructed to enable the i2 demand and supply planner applications to share.
Nike claims the kinks were ironed out by November 2000, but the damage was done, affecting sales and inventory deep into Nike’s next quarter. When the company’s SAP system arrived, short- and medium-range planning moved out of i2 altogether and into SAP. Nike says the $10 million i2 system was a small part of the $500 million overall project cost, although some observers assert that the i2 cost was higher.
Why did things go so wrong? Wolfram says Nike lulled itself into a false sense of security about the i2 installation because, by comparison with the SAP plan, it was a much smaller project. (Nike has about 200 planners who use the demand and supply planning systems.) “This felt like something we could do a little easier since it wasn’t changing everything else [in the business],” he says. “But it turned out it was very complicated.”
“Could we have taken more time with the rollout?” asks Steele. “Probably. Could we have done a better job with software quality? Sure. Could the planners have been better prepared to use the system before it went live? You can never train enough.”
Nike Learns Patience
Nike learned from its mistakes. There would be no rushing the SAP installation. And even though Nike executives occasionally questioned the project’s complexity and expense, Steele never considered abandoning the single-instance strategy. “We said single instance is a decision, not a discussion,” says Steele.
Nike wanted to do a staged, geographically based rollout of SAP, but it also wanted to avoid making each rollout so specific to a region that it would require specialized support. That meant building a design for the U.S. rollout that accommodated some of the peculiarities of the EMEA rollout—such as multiple currency support and different legal restrictions—even though those things were not required for doing business in the United States. This necessitated creating a global template for SAP processes, with all the regions agreeing on the minutiae of doing business. Naturally, this made each rollout longer and more complex.
Canada, a relatively small (roughly $300 million) piece of Nike’s $11 billion business, went first, on Thanksgiving weekend 2000 (the pre-spring rush quiet time), with SAP’s AFS ERP, a bundle of i2 applications and Siebel’s CRM system. Steele and regional Nike executives, dressed in smocks, served Thanksgiving dinner to project employees working around the clock. Other regions—the United States and EMEA—followed on successive Thanksgivings, putting 6,350 users worldwide on the system by the end of 2002. (The last two regions, Asia-Pacific and Latin America, are scheduled for rollout before the end of 2006, according to Nike.) Steele claims he’s never had to serve humble pie along with the turkey, saying to date there have been no disruptions to Nike’s business from the three rollouts.
This may be because of Nike’s newfound respect for training, another weakness of the i2 implementation. Nike’s U.S. customer service representatives received 140 to 180 hours of training from highly trained fellow Nike “super users,” says Andy Russell, Nike’s global transition director. Employees are locked out of the system until they complete the full training course, he says.
What Phil Knight ULTIMATELY Got for His Money
So what have six years and $500 million done for Nike’s business? Wolfram claims that better collaboration with Far East factories has reduced the amount of “pre-building” of shoes from 30 percent of Nike’s total manufacturing units to around 3 percent. The lead time for shoes, he asserts, has gone from nine months to six (in some periods of high demand, seven). But John Shanley, managing director with Wells Fargo, says, “Retailers are saying it’s still closer to nine months than six.” Gross margins have increased slightly since 2001 but not significantly.
Inventory levels have been reduced, says Supply Chain Vice President Dewey, by cutting Nike’s factory order interval time from one month to a week in some cases. But here, too, the effects may not be trickling down to the balance sheet as fast as Nike would like. Inventory levels are still at the mercy of Nike’s fickle audience of teens. Nike’s inventory turns were 4.34 per year in 2003, according to Footwear News, an industry trade magazine, slightly less than the industry average of 4.39 and behind rivals Reebok (5.07) and K-Swiss (4.47).
Nike also is behind its rivals in direct point-of-sale (POS) integration with retailers, says Shanley. Supply chain experts agree that actual data from stores, rather than software algorithms, are the best predictors of demand. But Nike’s SAP system cannot yet accept POS data, though the company says it’s working on it.
So far, the most direct benefits of the system have been typical for ERP: improved financial visibility, cash flow management, revenue forecasting, and an ability to juggle Nike’s cash stockpile in different currencies to take advantage of shifting exchange rates—benefits that are enhanced by the single database that holds all the data.
But Steele maintains that the best is yet to come. “We haven’t changed our processes too much yet,” he says, “because we didn’t want to complicate the rollouts.” Eventually, he believes Nike will get that six-month lead time down to three. But, he cautions, that that would require “significant changes on the part of our retail and supplier partners as well as Nike processes.”
He’d better hurry. Shanley says the sneaker market has changed a lot since Nike created its Futures program in the ’70s. Retailers don’t like having to order products six months in advance when fashions can change in a flash. Rivals are allowing retailers much more leeway in ordering practices, eroding Nike’s market lead in select areas.
But because Nike developed a plan in 1998, and stuck with it, the company claims it can make a coordinated global effort to cut that lead time. The system to make that happen is in place—which, given all that has transpired in the past seven years, is rather remarkable.