[From iSource Business, September 2001] In certain retail stores around the country, fans of Nike's Air Terra Humara 2 running shoe have hit the jackpot. Once selling for over $100, now these shoes can sometimes be picked up for less than half that amount. But don't thank Nike for that largesse. The cheaper shoes are by-products of the now infamous breakdown in Nike's supply chain a breakdown that CEO Philip Knight dramatically attributed to software supplier i2 when the company issued an earnings warning earlier this year. This is what we get for our $400 million? he fumed to financial analysts.
This refers to glitches in Nike's newly deployed i2 demand and supply planning application, which allegedly caused the athletic sporting goods conglomerate to overestimate demand for shoes in some geographies and underestimate demand in others.
The results were predictably disastrous. Some materials were over-purchased while the inventory levels of other materials dropped. The wrong shoes were over-manufactured; the most popular ones under-manufactured. To speed the right shoes to market, the company had to spend around $5 a pair in airfreight cost, as opposed to the usual cost of 75 cents by ocean. In all, Nike attributes some $100 million in lost sales to the debacle.
Finger-pointing ensued. i2 claims Nike refused to follow its advice to perform a standard implementation, instead forging ahead with an elaborate customization. Nike claims i2 over-promised its integration capabilities, vowing that, for $400 million, a little customization wouldn't be a problem. At least one industry watcher speculates that Nike is using i2 as a scapegoat for its poor earnings performance. Neither company is speaking about the matter in depth. Reputations have been tarnished. Lawyers have been summoned. The only winners appear to be the running enthusiasts who are snapping up Nike's Air Garnett III models for less than a song.
It Happens All the Time
Chances are, outsiders will never know exactly what happened, which is unfortunate, if only from an educational perspective. Demand forecasting and supply chain planning and execution are still, despite the commercial and technological advances of the last five years, incredibly difficult. Because it happened to Nike, we all heard about it, says Jeff Payne, CEO of Cigital, a software risk management consulting firm. But breakdowns like this happen in supply chains all the time.
But if Nike Nike! can't get it right, then who can?
The unfortunate truth is that too many companies, from the smallest mom-and-pop organization to the largest conglomerate, don't grasp the importance of supply chain optimization.
Payne cites such examples as W.W. Grainger, a distributor of B2B maintenance, repair and operation (MRO) supplies that lost $19 million in sales and $23 million in profits two years ago because it claims its SAP system over-counted warehouse inventory. Indeed, 1999 was a bad year for a number of companies. Hershey's third-quarter sales, which usually experience a boost from the candy-intensive Halloween holiday, dropped by more than $150 million, or 12 percent, from the previous year because of a glitch that prevented the sweets from being shipped on time. Additionally, Whirlpool had delayed product shipments because of such glitches, as did Snap-On Tools, which said it lost $50 million in sales when its inventory was miscounted. Nor has the problem gotten any better. In a survey conducted last year, Cigital estimates that companies lost half a trillion dollars due to similar software shortcomings.
It's probably more common for a company to spend $100 million and not get a strong return on investment (ROI) from the software, than to spend $100 million and get a lot out of it, says Rob Austin, a Cutter Business Technology Council Fellow and assistant professor at the Harvard Business School. There are a lot of those stories floating around.
This is not to say these breakdowns can always be traced straight back to a bumbling supplier although Nike's complaints of over-promised results did hit home with more than a few companies. Companies themselves often contribute to fissures in their supply chain structures. A groundbreaking study conducted by Deloitte Consulting a few years ago found that, for all the investments in supply chain technology, only a few companies had built truly integrated supply chain operations. Some 73 percent of North American manufacturers surveyed rated their supply chain as average or below, while only 2 percent ranked it as world class. Deloitte Consulting also discovered that some 47 percent of respondents had not developed a formal supply chain strategy, even though nine out of 10 rated supply chain management as critical to their company's future success.
A supply chain extends from the supplier's supplier to the customer's customer, says Marilyn Darling, a principal with Signet Consulting. That is a lot of hands that have to be willing to tow the line. You can't just send them a letter and tell them they're being supply chained.
Choosing the wrong application or addressing the wrong problem can also doom a project from the start. But perhaps the most oft-cited reason for failure is faulty implementation; indeed, system and business process integration has sucker-punched many an otherwise-savvy company.
The news isn't all dismal, though. There is a failure rate to consider, yes, but it's declining from where it once was, says Steve Gold, managing director of KPMG Consulting's Supply Chain Solutions practice. The tools are much more user friendly and their functionality has become much more sophisticated, he says.
Certainly the knowledge base surrounding this particular space has increased exponentially in just a few years. Ten years ago a company's supply chain might have developed haphazardly as new products were invented, factories were built or acquired, partnerships were formed, and fledgling markets were explored. Today's sophisticated supply chains, however, can resemble any number of proven strategies. A supply chain group might support multiple products or divisions, or it might be organized strictly by region, or by business unit. Or project teams might implement certain e-business initiatives in conjunction with central management. Or it might be some combination of the above.
But theory can only take you so far. Ultimately, the best description of an effective supply chain comes from illustration. As such, iSource Business decided to dissect two portions of two companys' supply chain operations to see how they work and, more importantly, what makes them effective.
Cases in Point
The first company is Kodak's Health Imaging division. A manufacturer of X-ray film and other medical products, the division was concerned about the disconnected communication between demand data and its supply and manufacturing activity. In particular, executives there were concerned that it was taking some two to three months to respond to a change in demand. Managing health care supply chains can be different from managing supply chains for other consumer products, because stock-outs at a hospital can compromise patient care, says Sean Willems, a principal at one of the suppliers that helped Kodak realign its system. The Kodak story also provides insight into the problems Nike was having in a similar situation albeit in a different industry but on a scale magnified thousands of times. Examining how carefully Kodak had to track its demand information for one customer in one division to find the weak spot paints an expressive picture of just how difficult the task was that Nike faced.
The other company is Covisint, L.L.C., the independent e-business exchange for the automotive industry developed by DaimlerChrysler, Ford, General Motors, Nissan and Renault. Formed more than a year ago, Covisint promised to be all things to all people as long as they were original equipment manufacturers (OEMs). Today, Covisint has redefined itself and broken down implementation of its once encompassing mission into manageable parts. Even better, suppliers are becoming enthusiastic about the money-saving possibilities. In fact, many analysts believe Covisint's recent launch of critical supply chain management applications will go further to assure the marketplace's ultimate success than any other act or announcement or appointment could do.
Making internal execution a priority shows Covisint is finally getting it right, says Kevin Prouty, research director of automotive strategies for AMR Research. Before, their attitude was, We will build it, and you will have to come.' Now they are saying we are offering you all these great tools for communication and collaboration...please come.' These days, Prouty says, suppliers are actually pressuring Covisint to develop its products faster. Again, a comparison with Nike is in order, although against the Covisint example its recent supply chain breakdown seems less acceptable. It is a totally different environment from the Nike/i2 situation, says KPMG's Gold. Covisint is building a new business from scratch and is a much more complex endeavor.
Now would be a good time to note that a year ago Nike could well have been picked as a candidate for this article, with its ability to fine-tune its demand forecasting in such a way that inventory could be maintained at the most profitable levels. Obviously we would have been wrong, although, to be fair, once it gets back on its feet, Nike has the potential to develop a formidable and competitive supply chain, especially with its related customer relationship management (CRM) and enterprise resource planning (ERP) initiatives. And we could well be proven wrong by our current choices. As Tim Thomas, director of process excellence for Kodak's Health Imaging division says, Supply chains need to undergo a continuous improvement process to stay competitive. Change is always impermanent.
So is improvement and, thankfully, so is failure.
That's why, some four years ago, Thomas began to focus on Kodak's demand-pull manufacturing processes. We were looking at moving our manufacturing planning processes away from make-to-forecast to make-to-order, he explains. We didn't want to build products unless there was really a demand for them based on the actual end user product usage. And once we built the product, we wanted to deliver it to the end user as quickly as possible.
It was at that point that Thomas and Willems noticed there were a lot of stages of inventory between Kodak, distributors, hospitals and the actual end users. They believed these stages of product and information batching could potentially hinder the overall efficiency of Kodak's integrated supply chain.
An independent operation in the overall supply chain might look good relative to its own specific goals, however, the total flow to the customer was very dysfunctional in terms of both cycle time and required inventory investments, says Willems, chief scientist at Optiant Inc., a provider of supply chain software.
So the pair began to lead a cross-functional team to reverse-engineer, so to speak, a supply chain with one large hospital group called the Health Alliance of Cincinnati, in order to discover if distortions existed and determine their root causes. We started looking at the channels of distribution, following the demand signal and trying to get as close as possible to the actual end user, says Thomas. We followed that signal through the hospital inventory, through our distributors inventory, through their ordering process and then to our own inventory and manufacturing planning processes. The confirmation: By the time the demand signal had reached Kodak manufacturing it was highly distorted. In other words Kodak was suffering from the bullwhip effect. The term derives from the idea that when you crack a whip all you have to do is move your wrist a little bit for the reach and angle to change significantly. There is a similar wave that moves through a supply chain, and each, slight variability is magnified the further that supply chain extends.
But the biggest challenge in determining what caused Kodak's bullwhip was a lack of product-specific information for the hospital's total inventory, as well as patterns of daily film usage. Similar information was lacking at the distributor, Marconi Medical Systems. And because Marconi represented several different hospitals, all with differing inventory management and ordering practices, the bullwhip effect was magnified even further once Marconi processed Health Alliance's demand data, as well as that of all its hospital customers, and then produced a demand signal to Kodak.
This batching issue for the information flow and product flow was compounded by inventory being replenished at both the hospital and the distributor level on a weekly, rather than a daily, basis. The root cause for this was the fact that Health Alliance was planning inventory levels on the basis of estimated usage or, in some cases, simply replenishing inventories according to standing orders, since the actual daily usage information was not available. In addition, individual operations within Health Alliance were responsible for their own inventory replenishment procedures using the multiple tiers of inventory at the Health Alliance sites, despite the fact that Marconi could deliver to any site within hours of order receipt, if necessary. If Health Alliance had been operating within an efficient integrated supply chain, it would have only required a week's worth of inventory on hand.
With this information on the table, Kodak worked with the hospital group and the distributor to minimize the number of hospital stocking locations and have the distributor deliver directly to the remaining locations. The hospital had many stages of inventory for radiology products in its different departments. We were able to get them to one fundamental stage and had our distributor work with them to reduce that amount of inventory to acceptable, but still safe, levels, Thomas says. Health Alliance also worked with Marconi to better monitor demand, actually eliminating the practice of standing orders, which was the cause of much of the distortion.
Since Health Alliance represented only one of many Kodak distributor locations, the specific Health Alliance process improvements were not scalable to fundamental process improvements and inventory reductions at Marconi and Kodak. However, by reducing stocking locations in the integrated supply chain, Kodak and Marconi reduced Health Alliance's inventory by 67 percent. Nevertheless there remains significant improvements to be had in supply chain efficiencies for both Kodak and the distributors.
Besides the grunt work of tracing the demand signal, Kodak also implemented a supply chain optimization application developed by Optiant called PowerChain Inventory, which focuses on determining optimal inventory levels and locations for a given customer delivery goal. I look at an integrated supply chain solution as having three legs to a stool, says Thomas. One is the inventory leg, and a second leg is the delivery performance to the customer. PowerChain helps us optimize both of these legs. The third leg is manufacturing technical, out-of-the-box quality. We are very focused not just on delivering the product in time but on the technical quality of the delivered product, as well.
Over the course of the project, Thomas didn't run into any unexpected barriers because [he] was expecting nothing but barriers. Some barriers were systems-information related, he says, and quite a few had to do with the people involved. For example, he says, while many of the hospital procurement staff was very approving of the project, some preferred business as usual. The same was true of the distributor.
The fact is, people will always have concerns when you implement a project of this sort, Thomas says. But if you can get them engaged in suggesting alternatives in a data-driven system and involve them in running the models, you can usually get them to come on board.
Now the Health Imaging division is using PowerChain Inventory to better manage demand in its worldwide equipment supply chain, with the goal of reducing two-thirds of its inventory. We want to reduce the echelons of inventory we maintain but still achieve our delivery goals to our customers, Thomas says. It's a big goal all right, but now that it's been accomplished on a micro level, there is no reason why Kodak can't take this accomplishment globally.
Half a year after it was announced, the automotive exchange with great fanfare finally unveiled its name: Covisint. The etymology was carefully explained in a press release. Co represents connectivity, collaboration and communication. Vis is the visibility that the Internet provides and the vision of the future of supply chain management. Int represented the integrated solutions the venture was promising to provide, as well as the international scope of the project. The name, said analyst Prouty at the time, was more a reflection of a lot of people wrangling for turf than anything else.
Clearly it was not an auspicious start.
Despite the somewhat ungainly name however, Covisint's goals were very visionary, especially for the times (a phrase that, yes, certainly sounds ridiculous considering this was only 18 months ago, but these are Internet months we are talking about). But the goal was to link more than 30,000 suppliers in one marketplace. Demand planning, inventory control, product design and overall supply chain management was to become much easier, since the exchange would allow the OEM auto manufacturers to view the progress of their suppliers' manufacturing operations.
Suppliers, however, were instantly wary, given the already huge leverage the automotive manufacturers enjoyed, and still do. Working together in a medium where they have insight into their operations would only make it easier for the OEMs to dictate pricing. Indeed, one particularly contentious issue was the transaction fee the OEMs said they would charge on procurement activities within the exchange, even those activities that were strictly supplier-to-supplier.
The situation has changed drastically since then. Covisint has gone from dominating the world to making their world a better place, says Prouty, who has never been shy to criticize the exchange.
This is far from saying that suppliers have become trusting as lambs. Trust is still our biggest hindrance, acknowledges Chris Steele, supply chain line of business lead for Covisint. We are asking our customers to share information with trading partners in a way they have not shared before, and that takes a leap of faith. Pricing for the transactions is also still an issue. (Reportedly, Covisint is still planning on charging a percentage of auction transactions. For the supply chain collaboration service, though, it will be more of a flat fee model. There is still some concern among users that Covisint is not being very flexible on pricing and could price itself out of the market, especially since FreeMarkets is such a good alternative, Prouty says.)
But an about-face has clearly been made, both by Covisint and by the suppliers. The key, says Prouty and other industry analysts, is that Covisint began focusing on solving some very serious communications issues in the industry. The auto industry has solid standards but they are all fragmented along OEM lines, says Prouty. However, with its recent information technology (IT) initiatives, Covisint has become a giant translator of sorts for the industry. And, along the way, it is promising to save suppliers some very big bucks indeed.
We tell suppliers they will see inventory reductions of at least 30 percent, says Steele. Premium freight costs should decrease between 50 and 90 percent, and administration costs should decline by 30 to 70 percent.
In some respects, Covisint is still the behemoth that it was originally intended to be. There are three main lines of business in Covisint, Steele says. Procurement, which includes the auctions and request for quotes (RFQs) capabilities; the product development and collaboration functions, which will allow suppliers and OEMs to share engineering tasks; and supply chain management, which includes planning and execution. Of all these capabilities, it is the supply chain execution on which Covisint has focused the most. We wanted to lead with applications like this so we could offer a return on investment (ROI) within a three month period, says Steele.
The beauty of the strategy is that it fills a hole in the auto industry, and it meets the needs of both OEMs and suppliers. OEMs need to extract inventory data out of the supplier's legacy system, says Chase Kushak, a member of Covisint's supply chain team. We can get that data in a very unobtrusive process and suggest ship amounts. At the same time, [suppliers] are able to see levels of inventory on the plant floor in real-time. Prior to Covisint, says Don Perchard, another member of the supply chain team, suppliers were forced to cover themselves by holding additional inventory. The new system has gotten great testimonials from both buyers and sellers.
The technology that is powering this new level of visibility is SupplyConnect, the exchange's communications hub that allows suppliers to share critical information, such as material releases, production schedules and shipping notices. The implementation is quick (about two to three weeks) and very straightforward, Steele says, taking approximately four hours to learn to use.
Another key application is the fulfillment product, provided through SupplySolution, which became operational in early July. The agreement reportedly gives Covisint selling rights to iSupply, although SupplySolution will market its products in other industries. The partnership with SupplySolution originated from requests by suppliers that wanted to see it implemented in Covisint, says Steele. We meet with our customer council every month so they can co-develop products with us. We know we have to please them and make it worth their while to participate.
Not quite. Automotive suppliers know participation in this exchange and others like it is inevitable if they want to continue to do business with the OEMs, who are, after all, the only game in town for them. But the advantages for the suppliers are becoming clear, as well.
SupplySolution's iSupply product has a successful track record of delivering immediate benefits to users, Prouty said in a recent report on the partnership. Automotive suppliers are struggling to maintain and improve cost positions during OEM production slowdown. Covisint has been slow in delivering on the promise of real value for suppliers, but now iSupply gives it the means by which to do so and quickly.
Learn and Live
Even among the Fortune 500, Kodak and Covisint are in their own league, making it hard for other companies to make apt comparisons and learn from their examples. However, it is not impossible. Indeed, many forward-thinking consultants and analysts believe the answer to supply chain breakdowns lies in a continual education, or knowledge-sharing process.
When Darling of Signet Consulting heard about Nike's problems she recognized from recent, first-hand experience many of the issues. The only difference was that her client which was undergoing an implementation of similar magnitude managed to avoid the worst of the problems by phasing in the implementation, requiring a great deal of collaboration among the groups as the project was rolled out. Not that she is feeling smug about dodging the bullet herself. In the last year companies have been rushing to implement their own e-business solutions because they felt they had to compete with the dot-coms and have competitive supply chains. In large part this was true for the times, she says. But it is how she believes Nike fell into trouble. Nike probably felt like they had to get everything out there, all at once. Now she says companies are not rushing to implement massive projects anymore, taking a wait-and-see attitude with the economy instead.
Her client, a multinational tech company, realized early in the deployment process that they couldn't afford to replicate mistakes as they increased the scale of their deployments. Companies need to front load the deployment process so they make all their mistakes up front and those mistakes don't snowball as the project is implemented site by site.
Front loading, Darling says, basically means a company consciously pays attention as it begins a project, looking for early indicators that something is not going right. Then you build in a lessons-learned processfrom the first deployment. It's a self-correcting process.
Many companies already incorporate lessons learned from their implementation process into a project, but they tend to do it at the end of the process, not the beginning. And that, I think, is not only ineffective but also cruel, she says. It is cruel because, by that point, it's too late for a company to use what it learned to make corrections and improve the outcome. Instead, it's an exercise in finger pointing.
Nonetheless, many projects are designed for implementation in just that way. A systems integration is most successful when you break the project into small pieces and go live with each piece, says Scott Moses, assistant professor of Industrial Engineering at the University of Oklahoma's College of Engineering. That way, he says, echoing Darling, you accelerate the learning, which in turn accelerates the implementation, ultimately making you more successful. Unfortunately, he says, many software projects are not implemented incrementally. Often, implementations encounter unexpectedcomplications if they are implemented as a single, large project.
Fortunately, it's the supply chain methodologies that can be easily broken down into bite-sized pieces.
Darling's client established an e-business learning methodology that pushed the various functions and geographic regions to have a voice in the implementation. Virtual knowledge-sharing conferences among all the deployment teams allowed the core design team to spot and address certain developing issues on the spot, including the customization issue that i2 claims derailed Nike to a large extent. Another problem the client managed to diffuse was a certain level of friction with one of its key suppliers. There was a corporate culture clash between the two organizations in terms of work style and information flows. Instead of escalating to a visible crisis, the client and supplier resolved their differences, she reports.
But even for those companies that do not wish to establish a formal virtual learning process on top of a complex integration project (although it certainly sounds like the wiser course), a certain introspection before and during implementation can mean the difference between failure and success, assuming you follow through with some intensive first aid. Darling suggests the following questions as a start:
- Are you making a big change in the way you do business, such as implementing an e-commerce, CRM, SCM, Six Sigma quality project; or just a plain vanilla integration with an acquisition?
- Are you starting something where failure would be decimating?
- Is there tension between two seemingly irreconcilable goals, such as standard implementation versus customization, speedy rollout versus customer relationships, or marketing versus engineering?
- Are the assigned tasks just overwhelming and your people feel swamped with no end in sight?
- Is employee turnover creating corporate amnesia causing you to reinvent the wheel as you go?
- Are your tried and true human resource tools no longer producing results?
- Does your project need to acquire a new tier of leaders because there is no sense of accountability, there is weak decision-making, or there is no clear succession plan?
An excessively expensive customization. An unprepared workforce for the changeover. Project creep (when an implementation grows beyond its original scope). Poor communications with a supplier and across the project. Customers and suppliers that don't buy in on the data and business process standards. These are the issues that can tank an otherwise well-intended, well-designed implementation process, Darling says. Identifying them early in the process is more than half the battle.
A battle that, in the rush to deployment, too many companies are losing.