At Nike, they call it being "outside the berm". Literally, they mean being outside of Nike's world headquarters in Beaverton, Oregon, where the company's lushly landscaped spread is encircled by a grass-covered slope that prevents outsiders from seeing in. But until recently, for Nike's overseas employees, being outside the berm also meant being out of the loop.
Inside the berm is where Nike's entire line of swoosh-bearing sneakers is born, hundreds of new shoes every season. If you were outside the berm, as Nike's Ellen Devlin learned when she took a post in Hong Kong three years ago, there were only two ways to stay plugged into Nike's latest designs: travelling on an aeroplane for the 15-plus hour flight to headquarters every three or four weeks or waiting for someone in Beaverton to ship the latest samples or photos across the Pacific. "If you went offshore," Devlin says, "you'd go dark." Now, thanks to its intranet, employees outside Nike's berm are no longer in the dark. Called the Global Product Information Network (GPIN, pronounced gee-pin), the intranet has become an online entryway into Nike's complex footwear production process. It's a central place where employees worldwide can find everything from the earliest sneaker sketches to the latest sneaker sales data.
Mike Caldwell, Nike's global director of brand knowledge creation and sharing, credits the intranet with improving communication between Nike headquarters in Beaverton and the international offices, which account for more than 40 per cent of the company's $US6 billion in global footwear sales. That's not bad for a project that cost Nike just about $US3 million to develop, notes Caldwell, a runner himself who has cross-trained in a variety of roles ranging from sales to IT since joining the company in 1983.
Shortening the Time-to-Sneaker
To understand GPIN's importance to Nike's footwear division (the apparel and equipment divisions do not use the tool), it helps to understand the cycle that turns a designer's dream into a shoe. Each footwear style takes 16 to 18 months to go from sketch to store shelf; at any given time, employees are juggling schedules for three seasons, each of which sees the debut of hundreds of new products. Even though Nike sells its sneakers around the globe, footwear design and development originates in Beaverton. The campus is home to footwear designers, who wallpaper their cubicles with the sources of their inspiration -- Superman comics, souped-up cars, sports stars. It also houses Nike's footwear developers -- those technical types who understand foot biophysics and high-tech sneaker materials, and create prototypes from designers' drawings -- and footwear marketers or product development managers.
In 1995, as Nike was about to begin a blistering three-year growth spurt that would boost its annual sales from $US3 billion to $US9 billion, it also began a footwear process re-engineering effort called Future Vision. "We asked, 'Hey, if we were going to do this thing from scratch, what would we do?'" Caldwell says. The Future Vision team came up with more than a dozen new concepts for how to make products, he says; for example, one concept was that of continuous merchandising, basically, looking at how a product fits within a merchandising scheme from the beginning rather than waiting for product samples. Each proposed improvement required a seamless flow of information to make it happen, Caldwell remembers.
At around the same time, Caldwell visited the Palo Alto, California-based Sun Microsystems to learn more about the latest Web technology. Back then, Sun was talking about giving each of its employees an intranet home page. Nike already had a traditional corporate intranet, called SwooshNet (after its famous logo), that offered company news and HR-related resources. Caldwell, who was then IT director for Nike's footwear division, thought giving each Nike employee a home page would be like giving each employee a way to waste time on creative efforts that customers would never see. But giving each Nike product a home page could help provide the seamless flow of information needed to make Future Vision a reality.
From that visit, Caldwell and his team decided to develop an intranet for Nike's footwear division to achieve this lofty goal: enable a seamless flow of information throughout the 18-month footwear development process and, in turn, improve global collaboration. What was their definition of seamless flow? Being able to enter and update data one time, in one place and share that data with any browser-using Nike employee around the world. The team had high hopes for the business benefits that such an intranet would offer. "We would be making smarter decisions, better product decisions and better products," says Brad Lawrence, Nike's manager of INet technologies. "[We would] be more of a global company, as opposed to a Beaverton-centric one." Web technology had several advantages over traditional client/-server development, Caldwell says. An intranet could tap into existing disparate legacy systems, such as Nike's Oracle-based product information system. It would also be less expensive to develop than a traditional application. "We could take some risks," Caldwell says, his rapid-fire delivery softened by a Southern twang. "If the sites didn't work, they weren't million-dollar sites, and the information was available somewhere else." Put Me In, Coach Caldwell's team members organised the first version of GPIN around job function: planning, design and product management. They wanted to include a financial section on it, but since Nike was in the middle of moving over to an SAP supply chain management system, they decided to wait to add the financial links. Caldwell's team set up an in-house studio for making digital photos of footwear prototypes, photos that could be posted on the intranet and updated whenever anyone made a change. That way, overseas merchandisers like Devlin would always have access to the latest product designs, without waiting for an air shipment or having to tote a digital camera to a meeting. Caldwell also decided to keep GPIN separate from SwooshNet, since GPIN was initially designed just for the footwear group. "People are very tribal," Caldwell says. "If this was for everybody, they wouldn't want to use it." The intranet debuted in June 1997 and was accessible to about 200 employees.
The GPIN team members thought employees would just love using the new tool, Lawrence says. They couldn't have been more wrong. Overseas employees like Devlin loved the idea of GPIN. But Beaverton-based employees didn't always update GPIN when product information or designs changed, making it hard for overseas employees to rely on it. The problem was that the Beaverton employees didn't need to use GPIN to do their jobs, and there was little incentive for them to put the extra effort into keeping it up to date. A marketer, for example, could run down the hall to grab the latest shoe prototype; she didn't need to go on the intranet to see it. A designer thought collaboration meant taking a sketch, putting it on his cube wall and having colleagues pop by to see it. And when he needed to collaborate with colleagues overseas, he felt like it was more secure to package up that sketch or design board and airship it to Asia than it was to post that sketch on the intranet.
To overcome that barrier, Caldwell assigned four information experts to be coaches; they spent the next nine months attending meetings, becoming familiar with how people used GPIN, explaining that GPIN was a secure way to communicate and prodding them to update it more frequently. "We had to retrain [them] that this is an important part of your business," Devlin explains. "If Japan doesn't know what's going on, it hurts your business." The coaches were also the eyes and ears of the GPIN developers. If they heard any complaints, they would ferry them back to the developers so the developers could quickly make improvements.
A second barrier to GPIN use was that individual product categories -- running, basketball and so on -- had already started putting up their own internal Web sites. Caldwell's team shut down the individual category intranet sites and convinced them to put their information on GPIN. They also started posting information on the intranet that even their Beaverton employees would not be able to get elsewhere.
Devlin, who by that time had moved from Hong Kong to Japan, could see the coaching start to make a difference. Asian category reps started showing up to meetings better prepared, toting printouts from GPIN showing the latest information on their product lines. There was still a difference category by category, however, as to how well their Beaverton counterparts kept GPIN up to date. "Some of them were very committed to their global team-mates," she says.
"Others were less connected."
A June 1998 GPIN redesign grouped content into category areas rather than job functions. Since employees generally work in only one category, it made sense for a running designer to be able to see information only related to running or for a tennis developer to see information related solely to tennis. The virtual realignment dovetailed with a Future Vision-inspired office space realignment in Beaverton that moved all employees in a particular category to the same location; previously, employees were grouped together by job function.
The latest version of GPIN offers rich product information online, from the very earliest stages of the design and planning process. Marketing managers can enter their earliest production forecasts, which in turn flow through into a corporate forecasting tool. Designers can scan their sketches and design boards and "check them in" to GPIN. The check-in form was designed to look like the dialogue box for creating an e-mail attachment, something that most GPIN users already knew how to do. Merchandisers can launch an application that lets them sort product lines by colour, price and target audience; the application itself is Windows-based, but it draws product information from the intranet. Colour designers can view future colour palettes and discuss which colours they think will work in which markets. Executives can view confidential reports on the top sellers for a given season. And perhaps the most popular area of the intranet is the virtual catalogue, which pulls information dynamically from disparate back-end systems into a single page. Someone who wants to see all of the men's running shoes, for example, will see a page showing photos of each style in all planned colours, plus information on each style's materials, target audience and retail price.
The number of employees with GPIN access has grown steadily, to more than 1500 at this writing. But for security reasons, Caldwell has had to limit GPIN access. Much of GPIN's product information is highly confidential when it is first posted but then becomes less so as the product moves through development and manufacturing. For example, the planned design of the hottest sneaker in the fall 2000 line should be seen only by a handful of people in January 1999; by the time the line launches, those product details can be shared with the sales force, customer service reps and retailers. Caldwell's team has not yet found an easy way to manage who has access to what information at different points in the product process -- say, to make those sneakers' details available to only a few people on GPIN at the start of the process but then available to thousands of people later in the process. Any solution to the life cycle security problem would need to involve workflow, Lawrence says; they would either have to write the software from scratch or heavily customise an off-the-shelf product. Another hold-up is that they want to use the same security infrastructure throughout the company. But Nike, like many companies, does not have a uniform product model -- that is, a consistent way of identifying products across its footwear, apparel and equipment lines. A running shoe, for example, is described by a number size; a running T-shirt is described by a letter, S, M, L or XL. Until the company comes up with a uniform product model, he says, it cannot roll out a life cycle security infrastructure companywide.
Has GPIN achieved its goals of improved global collaboration and seamless flow of information? Devlin, who returned to Beaverton in January, says it has. She finds GPIN to be an essential tool in her new role as product creation director for Nike's branded athletic wear. Her category works just 12 months ahead of the market, so it has only two months to make the kinds of product design decisions that other categories have six months to make. GPIN gives her a way to quickly show designs to her European and Asian partners. They have fewer meetings, and when they do meet, Devlin's overseas colleagues show up better prepared: they have already seen her line on GPIN, made suggestions for changes and maybe even seen those changes incorporated into the designs. That way, the meeting becomes a place to finalise plans and to finalise them at an earlier point in the product cycle. Since the front end of the process -- design and development -- is faster, there's more time in the back end of the process for wear testing, quality control and production planning. "We become a more disciplined and better company because of this tool at the front end," Devlin says.
GPIN's virtual catalogue feature has also streamlined access to Nike's cumbersome legacy systems. "I feel like we got a dinosaur to run the marathon and win," jokes Michele Young, Nike's footwear marketing operations manager.
Since overseas employees can view footwear designs online, they may not need to see as many physical shoe samples, Young says, which can also help cut development time and costs.
Those anecdotal testimonials are impressive. But Nike's CIO, Ken Harris, has pushed GPIN's team to put some hard numbers around those anecdotes -- that is, to measure how much GPIN has helped to reduce cycle time, to improve collaboration earlier in the process, to cut sample costs, and then to measure whether Nike has got any benefit from those changes. Harris, who left Nike in August to become CIO of The Gap, notes that Web applications are more iterative than traditional projects. In the old days, if a company wanted to build a new inventory system, it would come up with a list of specifications, build them out fully and then deploy the system. With Web applications, a company is more likely to build something quickly, measure whether it works, then decide to modify it or move on. "There's a bit of value creation based upon faith," Harris says.
At this stage, GPIN is used only by Nike's footwear division because it was developed by the footwear division. Under Nike's old IT organisation, each business unit had its own IT group, and there was little technology exchange among them. Since Harris has recently reorganised Nike's IT department into a centralised, shared-services organisation, it's much more likely that GPIN's technology infrastructure will be used someday by Nike's apparel and equipment divisions. The speed of its spread will depend on individual business units' IT priorities. Given that the infrastructure is already in place, other units' deployment costs will go down, and their cost-benefit ratios will increase.
Harris predicts GPIN will snowball.
More immediately, however, the GPIN team is planning another update. Nike lives by its production calendars, Lawrence says. The updated version of GPIN will have a calendar view, so users can click on events and publish or view information related to them. The new version of the intranet site will also have a product view that lets users choose products that they want to keep tabs on and select what information they need to know about them (for example, design boards, planning forecasts and colourways). That way, the information on a particular shoe will all be in one place on GPIN rather than scattered in several places.
Devlin says there's still a few improvements she'd like to see, such as having GPIN push new information to users. Yet as someone who suffered the information blackout of being outside the berm, she's thrilled with everything GPIN has delivered to date. "What we asked them to design was completely impossible," Devlin says. "What we got works better than we could have ever imagined it would."Swoosh School Intranet lessons from Nike's Global Product Information Network * Just do it. It's better to quickly develop an intranet application and then keep revamping it with new features or new links to back-end systems than it is to wait.
* If you build it, they may not come. Building an intranet is no guarantee that employees will use it. Without coaching or incentives to change their behaviour, employees won't bother.
* Put up the numbers. Anecdotal reports showing that people like an intranet help build credibility, but hard numbers are also a must to demonstrate value.