The China riptide is sweeping companies into a headlong rush to source from China — without thinking much about the trade-offs between lower product costs and higher overall costs and reduced profitability. A better approach — a way to turn the riptide threat into an opportunity — is to reduce the time and variability in the China-anchored supply chains. Doing this successfully will dramatically reduce costs, boost margins, and improve competitive advantage.

By George Stalk Jr.

Supply Chain Management Review

May 1, 2006

Surface freight from Asia to the West Coast of North America is growing at a rapid (and, by historical standards, explosive) rate. Port and surface-transport capacities, however, are not. North American ports and rail systems are beginning to choke, and delays and uncertainties are increasing. Freight demand on North America’s West Coast has been growing at a rate equivalent to one Port of Vancouver per year, and a rapid expansion of port and rail capacity would be difficult given political pressures and formidable environmental resistance. In fact, the situations we describe in this article are going to get far worse before they start getting better. Nor is the problem limited to North America, Europe will soon experience similar challenges.

The North American West Coast ports and rail systems are implementing changes, but these are not likely to meet the growing demand for many reasons. The most significant reason is that each of the North American participants in the China-anchored supply chains has a narrow view of its role and a limited notion of the end-to-end potential for improvement. The changes being made or contemplated reflect this narrow focus. They are incremental, and their effects will be swept aside by the bigger forces of the China riptide.

The supply chain bottleneck is beginning to affect the performance of manufacturing and retailing companies that rely on surface logistics to get their goods from Asia to the heartlands of North America. But few executives in retail or durable-goods companies understand the magnitude of the challenge. Even fewer are investing against the phenomenon to reduce costs, improve profitability, and create competitive advantage.

We strongly believe that by focusing on reducing time and variability in the China-anchored supply chains serving North America and Europe, companies can reduce their costs dramatically, improve their margins, and build competitive advantage. We believe that such performance improvements will dwarf the more conventional profit-improvement efforts now under way in most companies.

Retailers live or die by a very simple creed: Stock products that sell; do not stock products that do not sell. As North American companies source more of their goods from China, the risk of getting this wrong increases dramatically. The effects of getting it wrong show more in the company’s economic profits than in the accounting profits. Accounting profits capture the "generally accepted accounting practice" (GAAP) costs, revenues, and losses. Economic profits capture the hidden costs of lengthening supply chains: increased inventories, overproduction and underproduction, write-downs of excess inventories, and, most important, the lost margins from stockouts. In reality, accounting profits may be positive while economic profits are not.

Some executives may be tempted to stop reading here and pass this article to their logistics people. That would be a huge mistake. To turn the riptide threat into an opportunity, companies need to take cross-functional actions that come together at the very top of the company and require strategic choices, investments, and initiatives.

So what can companies do? They can retreat by bringing manufacturing home. They can sidestep the problem by building landside capability in ports that are not yet congested. Or they can be still more creative and competitive:

•They can aggressively manage their China-based supply chain, looking for ways to squeeze time from it that competitors haven’t identified or exploited.
•They can explore alternatives that will minimize adverse effects on the supply chain. These alternatives may include options — such as increased use of air freight — that appear costly but may actually result in lower overall expenditure by reducing hidden costs.
•They can invest in premiums and capabilities. Premiums are the extra payments required to get preferred treatment from ground, sea, and air shippers, port services, and other suppliers. Investments in capabilities enable companies to be better than competitors at managing their business in spite of the problems on the West Coast. These initiatives can include cross-docking, facilitated portside handling, and "track and trace" capabilities to keep boxes moving.

Companies that make creative investments like these will benefit enormously, change the dynamics of their industry, and open a competitive gap that will complement and enhance advantages founded on merchandising and store management. Many companies, however, are not dealing explicitly with the China riptide challenge because they are underestimating the importance of fast, effective supply chains. Most know that supply chain management is important. But rarely do executives think of supply chain investments as an outright source of competitive advantage. They underestimate the magnitude — and the impact on profitability — of the hidden costs of longer supply chains, reduced flexibility, and lost gross margin from missed sales and write-downs.

In their rush to source from China, many companies are blindly walking into a strategic trap. The trap is thinking that sourcing from China will result in lower product costs when, in reality, the supply chain dynamics will drive up overall costs and reduce profitability — thereby creating an opening for a competitor. Their only salvation is if all their competitors make the same mistake. But if they have competitors that do not source from China or that do focus on supply chain speed, their competitors will be competing with a different set of economics. The first company to see and correct the strategic error of sourcing from China without an appropriate investment in supply chain dynamics to minimize costs will seal the fate of its competitors.

The Direct, Indirect, and Hidden Costs of China Sourcing

Given the dynamics of the environment outlined above, how can companies be creative and competitive in their global sourcing decisions? The fundamental first step is to understand the direct, indirect, and hidden costs of China sourcing.

Companies that source manufactured goods in China do so because of the attractive costs. Yet unit product cost (UPC) is only one part of a very complex picture to consider in outsourcing. As supply chains lengthen, they incur direct, indirect, and hidden costs. Direct and indirect costs include shipping, nesting and de-nesting of containers at both ends of the ocean pipeline, inventory storage, handling, procurement, insurance, and overall financing.

Hidden costs are more difficult to identify and quantify. These are the costs that arise from the dynamics of the supply chain. As supply chains lengthen, they also take more time to move product. The longer a chain requires to move a product from point A to point B, the more difficult it is to manage without fluctuations, which create cost. The fluctuations arise as a change in demand at point B ripples back through the chain. Moreover, the flow of information about demand is delayed as it moves back through the chain, and, as a result, distortions arise and build. Typically a very small but unanticipated change in demand at point B can produce a change in demand at point A that is three to five times greater. This phenomenon is explained by system dynamics.

At most companies, management thinks of these system-dynamics effects as forecasting errors, and the greater the forecasting error, the further into the future managers seek to predict demand. But because forecasting errors increase with the magnitude of the delay, their efforts merely aggravate the system effects still further.

Forecasting errors create and drive all types of hidden costs. For one thing, variations in production loads based on widely fluctuating demand cause factories to be overloaded and underloaded. This results in inflated variable costs from overtime premiums and expedited material handling as well as in fixed costs that are not absorbed.

For another, additional inventory must be carried through the supply chain to facilitate fulfillment and cope with slack demand. And that additional inventory must sometimes be discounted, with a resulting loss of margin. Then there is also the lost gross margin from inventory that could have been sold if it had been available when customers wanted it.

Another hidden cost is the expense of flushing defective inventory from the supply chain. The longer the supply chain, the more time is lost identifying the cause of a quality problem, correcting it, and resetting production. At the same time, more defective goods are being fed into the chain during this period. Costs must be incurred to remove defective inventory with quality problems and to produce additional goods to meet back orders and current demand, further straining factories. Finally, more costs occur as companies physically replace defective inventory throughout the chain with new product.

As the container flow through West Coast ports slows down, further hidden costs are created, including the overtime costs that frequently arise from idled ships, unreliable portside loading and unloading, and overburdened rail-transport services. There are also the costs of rerouting shipments to other ports; in the case of the North American West Coast, shipments would be rerouted to the East Coast, and the Gulf of Mexico.

How significant can these costs be? One way to get a handle on them is to compare the economics of a typical domestic supply chain with those of one based in China. Such a comparison or simulation can become very complex very quickly. So to simplify it, we have assumed:

• A domestic unit price of $10.
• A domestic UPC of $4 for a gross margin of $6.
• A China UPC of $3 for a gross margin of $7.
• An operating margin of $4 for the domestic producer and $5 for the China outsourcer after a steady-state supply chain cost of $2 per unit for both the domestic and China chains. (This is a conservative estimate for the China chain since shipping-related costs are going to be higher than for the domestic chain.)
• An average steady-state demand of 1,000 units per week.
• A variable and fixed cost per unit based on this steady-state demand.

We start with a set of assumptions about manufacturing and logistics costs. (See Exhibit 1). These costs are composed of those that are fixed at a steady volume of 1,000 units per week and those that are purely variable by unit.

For the supply chains themselves, we assume some basic, fairly average operating parameters. (See Exhibit 2). For example, the domestic and nonintegrated supply chain, which we define below, has:

• Manufacturing and sales in the United States.
• An end-to-end cycle time of six weeks.
• A manufacturing time of three weeks.
• Plant inventory targets of two weeks.
• Transit time from factory to distribution center of two weeks.
• A distribution center inventory target of two weeks.
• Transit time from distribution center to retail of one week.
• Retail inventory targets of half a week.

The transit times may seem long at first glance, but we are trying to capture typical order-to-delivery times for these segments of the supply chain, which are always longer than shipping times themselves.

Inventory targets are set in supply chains to trigger orders and the markdowns needed to relieve excess inventories when they occur. In this simulation, when inventories exceed targets, the price of the excess inventory is slashed 20 percent and sales are assumed to be immediate. There are key omissions in this simulation that seriously affect the economics (in the negative). These omissions are all real-world phenomena and include the following:

• A network of manufacturing plants for subassemblies and the transit times between them (for instance, manufacturing partially in Puerto Rico for tax reasons or sourcing a key electronic part from the United States for assembly in Malaysia).
• Cross-shipping between distribution centers and retailers.
• Possible variations in transit times.
• The use of Mexico as a supply base, which, compared with China, is cost disadvantaged but time advantaged.

Demand is never steady-state — it always varies — but for the purposes of the simulation, we assume an average demand of 1,000 units per week with a variation about the mean of 300 units. We assume no growth or decline over time. Here again, we have left out some key factors that could have a seriously negative impact on the economics of this simulation. These include: the ramping up of new products and the ramping down of old ones; seasonal and promotional sales variations; growth or decline in demand; and mix shift, since we assume a single stock-keeping unit. To capture the effects of omissions, we would need to construct a much more complicated simulation.

Our simulation also assumes three states of supply chain sophistication depending on how information on customer demand flows. Exhibits 3A and 3B show these information flows as well as the flow of a hypothetical perfect forecast supply chain.

• In a nonintegrated supply chain, each upstream step gets its information on demand from its customer on the next step down.
• In a semi-integrated supply chain, each step gets its information on customer demand from two steps down the chain.
• In an integrated supply chain, each step has a full view of final customer demand. For internal analysis purposes, we also include a representation of an integrated supply chain if it could have access to perfect knowledge of future demand. This is different from the integrated, semi-integrated, and integrated supply chains, which are based on historical averages of sales and product demands.
As Exhibit 2 shows, the simulation will also look at the effect of varying cycle times for both domestic and China-based supply chains. The simulation is demanding. There are four supply chain models, five minimum-order-quantity scenarios, and four cycle-time scenarios, resulting in 80 supply chain simulations. These models are subjected to normal random demand and are run over 3,000 weeks to escape the effects of initial conditions, where everything — inventories, production loads, sales, and stock balances — is by necessity in a steady state.

In the first set of simulations, these levels of information management sophistication are applied to a wholly domestic supply chain and to a chain with China at one end and, say, Chicago at the other. In our first comparison, the domestic chain has an overall cycle time of six weeks and sets its production run at one month of average demand. The China chain has an overall cycle time of 11 weeks. The China supply chain also sets its production run at one month of average demand. Information flows for both chains are nonintegrated, and weekly demand is allowed to fluctuate randomly about average weekly demand.

When the two supply chains are subjected to this demand profile, retail stock fluctuates between being overstocked and out-of-stock. An example of the system response to these demands is shown for illustrative purposes in Exhibit 4. The exhibit depicts how the four China-anchored supply-chain models, from nonintegrated to "perfect forecast," respond to random demand (the solid line) and to the resulting sales (demand less stockouts plus discounted sales—the dotted line). When overstocked, we discount the excess inventory in order to move it. When understocked, we count as a cost of doing business the lost opportunity of not obtaining the gross margin from the goods that weren’t available. The margin realized from sales revenues less supply chain costs is shown in green. The margin lost owing to stockouts is in yellow.

The results of the supply chain simulations: The domestic chain experiences volatility, which adds supply chain costs. The operating margin averages $0.77 as opposed to our assumed steady-state operating margin of $4. The China-based chain experiences even greater volatility because of its longer cycle times and nonintegrated information flows. Even with a lower UPC, it ends up with an operating margin of $1.02 as opposed to the assumed steady-state operating margin of $5. Because of its lower UPC, the China-based chain still has the advantage.

Competitive Implications of Lengthening Supply Chains

Exhibit 5 shows a series of scenarios comparing a China-anchored supply chain with a wholly domestic chain. The China chain is likely to move quickly to a semi-integrated state of sophistication as the difficulties of managing the nonintegrated chain become apparent. In doing so, it improves its operating margin to $1.21 per unit. At that point, the management of the China chain is likely to cut price by, say, a dollar per unit, to gain share from the domestic chain. It believes that the move will reduce its gross margin from $7 to $6 and its operating margin from $5 to $4.

If the domestic supply chain matches the price cut to hold share, it will see its operating margin turn into a loss of $0.16 per unit, while the China-based chain’s operating margin will decline to $0.28 but still be profitable. The reality is likely to be somewhere in between. But the China-based chain will still have the advantage through a combination of low UPC and semi-integrated information flows.

But now our scenario takes an interesting turn. The domestic supply chain can neutralize the China-based supply chain’s advantage if it integrates its information flows and cuts end-to-end cycle times by half to three weeks (still a big if for many companies). With that enhanced responsiveness, the domestic chain will see its operating margin increase from a loss of $0.16 per unit to a profit of $2.19 per unit. Now it has the advantage. However, the competitive dynamic might continue with the China-based chain becoming integrated and also cutting its cycle time in half. In that case, the advantage returns to the China-based chain because of its lower UPC.

Unfortunately, the world we live in isn’t evolving in a way that would permit these improvements for the China-based chain. As noted, the surface freight situation in North America and Europe is seriously challenged. Backlogs at ports and on railroads are at all-time highs. With freight volumes increasing faster than the ports can handle them, the situation will only worsen. Some Asian ships are too big to go through the Panama Canal to less busy ports on the Atlantic coast. Even some of those that can fit through the canal must offload and reload containers to meet the canal’s pilot-visibility rules. The offloaded containers are sent by rail across the isthmus to be reloaded on the other side, adding even more transit time. And while shifting to East Coast ports might improve the predictability of shipping times, it certainly won’t shorten them.

Because of the problems on the North American West Coast and in Europe, the cycle times of the China supply chains are going up, not down. If they increased from 11 weeks to 18, the China-based chains would suffer a decline in operating margin to a $0.70 loss per unit, while the enhanced domestic chains would still be realizing a profit of $2.19 per unit.

But that’s not all. The cycle times of surface shipments (from China to Chicago, for example) are not only increasing — they are also becoming more variable. About 50 percent of the containers at one shipping company are offloaded within two weeks of their scheduled dates, and these are considered to be on time. The other 50 percent are even less predictable!

If the surface time of 18 weeks can randomly vary six weeks either way, the semi-integrated China chain will lose $1.43 of operating margin. So a domestic supply chain with integrated information flows and fast cycle times can outperform a China chain, despite China’s low UPC.

In conclusion, the domestic supply chain should become time-advantaged with or without the threat of a China-anchored supply chain. The potential improvements to its competitiveness and profitability are too great for companies to ignore, although they still do.

What Companies Should Be Doing

Everyone rushing to source from China can’t be wrong. However, a growing number of companies are having second thoughts about their China sourcing strategies and are either reworking their North American logistics networks or retreating from China. Some examples:

• Wal-Mart has established a 1.3-million-square-foot warehouse near the Port of Savannah and will open an $80 million distribution center near Houston.
• The Home Depot has built a 1.4-million-square-foot facility in Savannah.
• Toyota and Nissan are considering a plan to move inbound logistics through Mexican ports.
• Red Bull has diverted all inbound shipments from Southern to Northern California ports and is redistributing products to local markets in Southern California as needed.
• Windbrella has eliminated its Florida warehouse and has enlisted UPS to ship umbrellas directly from its China production plant to U.S. retail outlets.
• Red Wing Shoes is having its products consolidated by UPS in China prior to shipment.
• Kodiak is bringing boot manufacturing back to North America.

So what can companies do? First, they need to be very aggressive in managing their China-based supply chains, looking for ways to squeeze time from them that competitors haven’t identified. For companies that haven’t yet sourced from China, we recommend the following steps:

• Reduce minimum-production-order quantities and reduce cycle times as quickly and as much as possible.
• Don’t source or manufacture in China until management fully understands the dynamics of the supply chains.
• Create an integrated or a semi-integrated information flow within the company’s existing supply chain.
• Conduct in-depth examinations of buying practices and supplier relationships management at all levels of the supply chain. Then identify and prevent potential hidden costs.
• Segment the demand chain on the basis of order predictability and demand volatility so that components with the highest gross margins and the most volatile demand get the fastest handling.

If a company does decide to source from or manufacture in China, it should explore alternatives that will minimize adverse supply chain effects. These alternatives may include options that appear costly but actually result in overall lower costs. For example:

• Using air freight for products with the highest margins and volatility.
• Insisting on point-to-point ocean shipping. To reduce costs, shipping companies build larger and larger container carriers, which must then be scheduled to call on multiple ports. Shipping products on a vessel that has your destination as its last port of call can add weeks — and great variability — to transit times.
• Developing better relationships with transportation providers. This could mean paying the shipping company for preferential treatment. In "hot hatching," for example, you offer a premium to a shipping company that will load your goods onto its vessel last and unload them first. Another option is to work with the few shipping companies able to offload containers directly onto rail cars that are then expressed east — cutting days and sometimes weeks out of the supply chain.

All these initiatives require investment in either premiums or capabilities. Premiums are the extra payments required to get substantially enhanced performance and preferred treatment from suppliers like port services and ground, sea, and air shippers. We have talked with many of these companies, and they do not know what to offer at what price. Companies can get results by forcing suppliers to compete on service in return for premiums.

Capabilities investments, which tend to be a good deal harder to discern and carry out, include such things as accelerating the flows and interpretation of information; developing designs that enable final assembly to take place close to the point of final demand, thereby minimizing the time and cost effects of long supply chains; learning to source, manufacture, launch, and withdraw products more effectively; and exploiting new concepts for fast freight.

Identifying these investment opportunities requires exhaustive investigation and analysis of costs, revenues, and lost margins in today’s end-to-end supply chain. Companies need to ask "what if?" and they need to probe their answers deeply before they decide that any additional investments in premiums and capabilities are not likely to produce further improvement. They need to be especially alert to the subtle but important system effects of investing in one part of the chain to affect performance at another.

Next Steps

Information is power, and companies need to invest in it. Here are six information-related steps that managers can take to improve a company’s supply chain-whether they are sourcing domestically, in China, or in some other part of the world.

1. Estimate the size of the prize. How will the ideas discussed here work in your environment? What are your special situations? High volatility? Fast fashion cycles? Customization? Distributed manufacturing?
2. Walk the line. Understand what is actually happening in the supply chain and why, step by step.
3. Focus on dramatically improving the responsiveness and reliability of key participants in the supply chain. Sometimes simple procedural changes can have huge implications.
4. Identify and vet the changes necessary within the organization and across the supply chain to realize opportunities. Companies are seldom organized to make the cross-functional changes required to create a material impact on the performance of a supply chain. One company’s purchasing group sourced the parts for a particular design from suppliers in three different countries solely on the basis of UPC and without regard for the impact of the decision on the entire system. The result was frequent assembly shortages and emergency air-freight charges.
5. Get information to the top of the company. Push efforts onto the A-priority list of the company’s leadership—or else abandon the efforts altogether. Dramatically improved supply chain performance has to have a mandate from leadership. Otherwise the near-term and narrowly focused performance parameters of the organization will bring all sustained effort to a halt.
6. Build improvement efforts into operating plans and budgets throughout the company. Some organizations simply build in stretch performance goals and let their organizations rush to find solutions. This method is scary to watch, but it can be very effective.

In general — and of utmost importance — the strategy must match the supply chain. If a company decides not to source goods in China while its competitor does, the rival’s direct cost advantage can sometimes be overridden by heightening its logistical disadvantage. For example, the company with the domestic supply chain could try to increase the degree of customization its customers want or to raise the fashion quotient by creating more variety or more frequent selling seasons in some category of its business. This would increase the demand volatility for certain products, and the China-anchored competitor, with its long lead times, could find its logistical problems aggravated.

Companies with time-advantaged supply chains might also consider consignment pricing, requiring their customers to pay only when they sell the company’s products. To match this appealing offer to customers, competitors will have to incur much higher costs for carrying greater inventory in a much longer supply chain.