Toyota, a half century ago, pioneered the just-in-time strategy for supply chains—and was celebrated for its ingenuity and efficiency. With just-in-time processes in place, companies match supply with demand, producing and receiving goods only as needed, which reduces inventory costs and waste.

However, the past two years have made clear the risks involved in such streamlined operations, and companies using just-in-time may be vulnerable to even small disruptions. And while Toyota and others have maneuvered through large disruptions—such as the 2011 earthquake and tsunami in Japan, which caused the carmaker’s production to plummet by over 60 percent in a matter of weeks—they’ve been humbled, as almost all producers have been, by the prolonged global disruption wrought by COVID-19. (See “How COVID-19 broke the supply chain,” below.) Consumers now experience delivery delays for a broad variety of products, from cars to couches to consumer electronics. Stores are subject to shortages and rising prices. Cargo ships are stuck at ports, waiting to unload.

The COVID-19 supply-chain crisis, says Chicago Booth’s René Caldentey, has raised red flags about how a relentless focus on efficiency has compromised reliability, and it has highlighted how prone the chain is to disruption. Now many companies are moving to address that. Even Toyota has been pivoting away from just-in-time and in September reportedly instructed some of its suppliers to store five months’ worth of semiconductors, rather than the usual three.

But just-in-time is only one piece of what has created today’s supply-chain puzzle. Companies also have to plan around the explosion in e-commerce, and an increasing need for sustainability, among other factors. “This is a very dynamic system that keeps evolving,” says Caldentey, who warns that when the current crisis is resolved, other shocks will emerge, including some related to climate change. “This is going to keep happening,” he says. “COVID has this compounding effect, but many of these problems [we’re experiencing] are not due to COVID.”

As experts look ahead to a postpandemic supply chain, they’re considering the big questions beyond crisis management, some of which get to the core of what supply chains are even about. What do we want from supply chains, and what do we want most? Should companies focus on delivering same day or on being more environmentally friendly? Should a chain be more resilient and able to withstand disasters, or should it be efficient above all else? Should it be shorter? Closer to home? More flexible? Can it be all of those things at once?

Make it transparent

Many supply chains are global and complex beyond our current understanding. Especially since China joined the World Trade Organization in 2001, companies have done deals with suppliers and producers all over, and they’ve come to rely on “long supply chains that snake around the globe,” according to the description of a National Bureau of Economic Research project called “The Rise of Global Supply Chains, Networks, and the Rise of Interruptions: Looking Forward past COVID-19.” The effort, led by Harvard’s Laura Alfaro and Booth’s Chad Syverson, seeks to measure the consequences of longer supply chains and consider strategies to reduce risk. Long and lean supply chains can be, and have been, interrupted by natural disasters, cyberattacks, and political upheaval.

“A lot rides on the existence and type of substitutes available within supply chains,” says Syverson. “How many companies supply a given input? How many countries and geographic locations can the input be sourced from? How costly is it for inventories to be used to insure against disruptions? Can air transport be used to substitute for ships? Basically, if interruptions occur, how costly is it to work around them? And does it make sense for companies to spend resources to expand such options before any problems arise?”

But before companies can make significant changes to their supply-chain processes, they have to get a handle on whom they actually do business with. “In the supply chain, we have six megaprocesses: plan, buy, make, move, distribute, and sell,” says Texas A&M’s Eleftherios Iakovou. Companies work with others to accomplish the six processes, but few know all the parties with which they are directly or indirectly doing business, and few have enough information to know where the next crisis could come from, much less how to address it.

How COVID-19 broke the supply chain

When an earthquake and tsunami hit Japan in February 2011, it caused supply-chain issues worldwide. In the auto industry, Fuji Heavy Industries, Honda, Nissan, and Toyota plants in the disaster zone were forced to close. Other Japanese plants further from the disaster halted production due to parts shortages caused, among other factors, by damage to their supplier plants.

This has been a problem since long before the pandemic. The lack of transparency has enabled labor exploitation, environmental destruction, and reputational damage. Greenpeace and other activist groups have called out multinationals for using suppliers linked to problems that have ranged from dumping toxins in rivers to committing labor violations. Without full transparency, including regular audits, companies can claim they weren’t aware of the less ethical aspects of their supply chain.

Booth’s Nicole DeHoratius says that the first step toward transparency is for companies to map their networks fully from tier-one suppliers down, so they understand the exact composition of their supply chain.

Not all companies have a strong incentive to be more transparent. Each individual link in a chain has different goals and makes decisions in its own best interests. This is a major obstacle to sharing data within the supply chain. Consider an automaker and a brakes supplier three tiers down on the chain. The brakes supplier could be concerned that if it were to share capacity or cost information, the automaker might try to squeeze costs, or even try to bypass it and work with the supplier’s own suppliers. “There’s just a total lack of trust on how this information would be used,” DeHoratius says.

Survey responses in the Business Continuity Institute’s 2019 Supply Chain Resilience Report indicate that 57 percent of global companies surveyed likely lacked complete visibility into their supply chain before the pandemic, and another 20 percent were unsure if they had it.

However, in light of the pandemic, “the urgency is there,” DeHoratius says. “The incentive conflicts haven’t gone away, but there is more talk about the necessity of developing this map and understanding the supply chain.”

The pandemic in perspective

Various events over the past 20 years have stressed supply chains, but none as much as the COVID-19 pandemic.

This talk involves the need for control towers, data centers that provide real-time metrics across the supply chain. These data can be used to identify potential disruptions, suggest alternative sources, and track product flow, among other things. But DeHoratius says the effectiveness of control towers depends on the integrity of the data they utilize, and her prior work indicates that data accuracy continues to be a problem. Overall, she notes an “accumulation of errors” across the chain that include inaccurate purchase orders, port arrival data, and item descriptions for customs clearance.

Iakovou concurs that a lack of good information sharing has hindered the ability to detect problems ahead of time and respond to crises faster. Indeed, visibility, transparency, and mapping are a few of the tenets for developing resilient supply chains. His research with International Hellenic University’s Dimitrios Bechtsis, University of Cambridge’s Naoum Tsolakis, and Aristotle University of Thessaloniki’s Dimitrios Vlachos outlines a three-layer framework for supply-chain managers to improve data, integrity, and flows (of products, information, cash, and processes) across networks.

In their framework, at the cloud layer, several data streams relevant to a company’s supply chain are collected, managed, and shared across the chain according to various confidentiality levels. A Blockchain layer stores all these data securely, and the A.I. layer analyzes and processes data from the Blockchain layer to create algorithms that help supply-chain managers make decisions.

Blockchain can be used to improve transparency by streamlining transactions, notes Auburn’s Glenn Richey. He says the technology helps companies to make some data easily accessible and to keep them updated, which improves internal forecasting.

And once good data are available and accessible, they also can be shared selectively. Research that DeHoratius conducted with the Ohio State’s Elliot Bendoly and Nathan Craig provides an example of how that could help, arguing that companies can reduce the cost of uncertainty in crisis times by tracking two specific metrics: consistency (the ability of a supplier to fulfill orders repeatedly) and recovery (the ability of the supplier to fulfill orders after a service lapse). These two concepts, says DeHoratius, “have been critical in today’s dialogue about managing the pandemic disruptions to the supply chain.”

Make it resilient

Opacity can hide risks in the chain. Before Japan’s devastating earthquake and tsunami in 2011, Toyota had close contacts with its tier-one suppliers and assumed that they were using a large variety of tier-two suppliers, DeHoratius says. After the disaster, Toyota took a close look at its tier-one suppliers and found that most were themselves using the same suppliers. The diversification of risk Toyota assumed was there actually did not exist, she says.

Iakovou argues that improving digital approaches to mapping extended supply networks is crucial. Then companies can make changes so they’re able to pivot more easily to other suppliers if needed. “Qualifying and engaging multiple suppliers of course has a cost, but, in times of disruption, those costs can be paid off multiple times over,” says Booth’s John R. Birge.

Having multiple suppliers can help both reduce the costs associated with finding and onboarding a new supplier in the midst of a crisis and avoid expensive production delays. “If more companies react to current conditions by increasing redundancy in their supplier base, that should lead to more resilience in the overall supply-chain network,” says Birge.

But building in redundancy costs more. And how much redundancy is enough? That requires accurately assessing the risk of a shock to the chain. What’s the likelihood of a tornado, energy-price shock, or pandemic that would make it worth the price of carrying extra inventory or having backup production capacity?

And what are the costs associated with assessing the risks, or not doing so? Before COVID-19, the World Health Organization put the risk of the spread of infectious diseases as below-average likelihood but above-average impact. Many companies failed to prepare, in part because of potential pushback from shareholders, and are reassessing how much it’s worth to be ready for a relatively unlikely event.

Recommended Reading

Make it flexible

Having multiple suppliers helps build resiliency, and so does flexibility, which Booth’s Yuan Zhong argues can pay off even in noncrisis times. Zhong, who studies flexibility in production plants, gives the example of a company that makes two types of cars, each in a different plant. The demand for each type of car changes over time, even over the short run.

What if one factory makes both types of cars and can adapt with demand? This same concept of flexibility applies to factories repurposing their manufacturing processes to make completely different products—just as when General Motors repurposed its factory in Warren, Michigan, to produce face masks during a critical shortage.

Of course, if you have a company that makes 100 types of cars, creating a plant that can produce them all is not cost-effective. It is usually more efficient to build a set of factories that can make three or four types of cars each, Zhong says—but then, which factories should make which cars? The key to making this work is robust demand forecasting so companies can plan years ahead.

“With a little bit of additional investment in your flexibility of capacity, you can better match supply with demand,” he says. “Then when you have a supply chain, and we’re thinking of matching supply and demand at every step of the way, this can help.”

Research by MIT’s David Simchi-Levi, Georgia Tech’s He Wang, and Duke’s Yehua Wei finds that when a production plant can build more than one product at the same time, this can help mitigate unexpected supply shocks. A company capable of adjusting its production capabilities in the case of disaster stands a better chance of managing supply and demand challenges, they conclude.

Just-in-time manufacturing sneers at the idea of keeping spare parts around, but Simchi-Levi, Wang, and Wei’s research suggests that holding excess inventory can make supply chains more resilient. The researchers imagine a case in which a company has different plants producing different products—let’s say one makes baby bottles and another makes wooden toys—but has the flexibility to change if needed. When a storm hits the toy plant, the company could use the extra baby bottles it has stored to satisfy demand while it rejiggers the bottle plant to produce toys.

“When a firm has both inventory and process flexibility, inventory helps to free up excess flexible capacities during an unforeseen event, and thus improves the firm’s ability to mitigate risk,” the researchers write, noting that it is generally more cost-effective to create plants that can produce more than one product, but not necessarily all of a company’s products.

The same idea also pertains to assemble-to-order systems, which involve keeping an inventory of product components that companies can turn into a range of products as consumer demand waxes and wanes. “This type of modularity is one way that supply chains can look to become more resilient amid the uncertainty of a post-COVID-19 world,” says Booth’s Levi DeValve.

Building resilience in inventory and component storage can also reduce kinks in the supply chain. Think of Amazon filling orders that involve items stored at multiple warehouses. But more flexibility also means more complexity, in which case efficient algorithms and similar optimization tools can help, he adds.

“It has long been known that flexibility in these systems has been an important tool for dealing with uncertainty, both from routine demand fluctuations or shocks like those from the recent pandemic,” DeValve says. “I expect this to become even more of a focus moving forward.”

The business case for sustainable supply chains

The push to create greener supply chains is picking up steam, as is the pressure to be mindful of social issues including fair labor, human rights, and overall ethical business practices.

Move it closer

Supply-chain managers are also reassessing where factories and partners are located. The COVID-19 crisis reinvigorated the discussion about reshoring—in the case of US companies, bringing manufacturing jobs back to the United States—by highlighting that doing so could help mitigate risk. But bringing all factories back, or building solely US plants (onshoring) is unrealistic, says Iakovou.

“Complete reshoring of supply chains cannot be the answer, as it would make US businesses less competitive and put them at a disadvantage with businesses of often adversarial nations, such as SOEs, the state-owned enterprises that dominate China’s economy and supply chains,” he says. To this effect, Iakovou, in an article he coauthored with Chip White of Georgia Tech and published at Brookings’s TechStream, is proposing the development of next-generation, cost-competitive, resilient supply chains that are based on data monetization and new manufacturing technologies.

A better solution than reshoring, say many experts, is to concentrate on ally shoring, or nearshoring. This involves identifying which plants are most conducive to moving back home or closer to home (say, to Canada and Mexico for US companies), then focusing on strengthening supply chains with these nearer regions, which offer better relations and more overall transparency.

This needn’t mean completely abandoning far-flung manufacturing locations, but it would still require a big shift in many supply chains. Even in 2020, while the COVID-19 pandemic raged, China was the US’s top trading partner by import value, with Mexico and Canada following.

Location, location, location

There’s value in having suppliers in different areas. While US companies with Chinese suppliers and customers saw borrowing costs rise at the start of the pandemic, their costs plummeted when Chinese factories reopened and the rest of the world locked down.

“There is a real opportunity for Mexico, Central America, and the northern part of South America to get some investment and become support suppliers across a number of different industries and then, over time, to become some of the main players for a number of companies,” Auburn’s Richey says.

That said, there remains value in doing business in distant countries, despite the political difficulties involved in some places, says Booth’s Birge. He explains that just as having multiple suppliers lessens risk, having suppliers in diverse locations can help insulate a company from shocks.

Research by George Washington University’s Şenay Ağca, Birge, Chinese University of Hong Kong PhD student Zi’ang Wang, and CUHK’s Jing Wu finds that when only China had locked down and shuttered factories in the early days of the pandemic, North American companies with Chinese partners experienced higher borrowing costs. But once China reopened and the rest of the world scrambled to prevent COVID-19’s spread, those companies saw their credit risk fall. The positive impact of those connections to China illustrates the value of having suppliers in many regions, as well as the benefit of not entirely reshoring or cutting ties with companies in countries such as China, says Birge.

As Iakovou notes, the US now has an opportunity to work collectively with its trading partners around the globe with a mix of offshoring, nearshoring, and reshoring. This strategy, he says, “will promote a new generation of diversified, resilient, and competitive supply chains, supporting employment, trade, and US prosperity, while further supporting the nation’s geopolitical prowess.”

Cut costs from the ‘middle mile’

In e-commerce, the final legs of the delivery process can account for up to half of the cost of shipping, due to complications ranging from traffic and accidents to failed deliveries and lost drivers. Considering the cost, companies work with researchers on how to improve the efficiency of these final trips, including both middle-mile deliveries, which move products from regional warehouses to delivery stations, as well as last-mile deliveries, which carry products on their final journey to the customer.

Make it work for individual customers

Part of what’s challenging about the supply chain is that it’s constantly evolving. Customers, for example, have grown used to shopping online and receiving immediate deliveries.

Responding to this trend, companies have gone on a warehouse-building spree. Amazon currently has more than 110 fulfillment centers in the US, up from just six in 2005, and plans to open more. Urgent orders must be fulfilled locally, which affects what retailers house where and limits their ability to, say, ship a product from a California warehouse to a customer in New York, or respond most efficiently to an order containing multiple products. As these fulfillment decisions grow increasingly complex and expensive, research suggests that better algorithms can help control costs and maintain efficiency.

But another view is that companies should rethink immediate delivery and retrain customers accordingly. “I think companies have, in an effort to try to compete on this metric, created bad habits in consumers,” says Caldentey. “People want things the same day and have gotten used to it.”

These same people may also value other things, such as information about when a package will arrive and flexibility in delivery time. In some cases, customers prefer precision and flexibility to speed, according to research by DeHoratius, University of Porto’s Pedro Amorim and Sara Martins, and Stockholm University’s Fredrik Eng-Larsson. The average customer, they find, is willing to wait almost 11 hours longer for a product if the delivery window is just an hour shorter, and will wait an additional 7.5 hours if the delivery will arrive on a preferred day of the week. Additionally, repeat customers may be willing to pay more for delivery than occasional customers, and someone making a large order may be willing to pay double the fee to narrow the delivery window by just one hour.

A waiting game

Companies have gone on a warehouse-building spree in order to deliver products as quickly as possible. But research suggests that customers are willing to wait longer for a delivery in exchange for a more precise delivery time.

Customers have also come to expect an omnichannel experience, in which they can deal concurrently with a company’s internet and physical stores. A busy parent buying a child’s rain jacket online might have it shipped home, but might also want the option of returning the item to a store, where the child can try on another size if the first doesn’t fit. The pandemic accelerated the trend, says Chicago Booth’s Linwei Xin.

And it may be impossible to reverse. Xin notes that some retailers, such as grocery stores, had to restructure their operations during the pandemic to maintain their business. That meant transforming their stores into small warehouses and offering delivery or curbside-pickup services. Walmart expanded its online business, and Amazon added physical stores, in addition to those it gained when it acquired Whole Foods in 2017.

Richey says the pressure to deliver an omnichannel experience can make companies more efficient. However, integrating online and offline offerings can be challenging. Consider a shopper who places an order at for pickup. A store associate runs around the store shopping for that customer.

“This is an additional step that is quite costly, because picking is labor heavy and is usually a bottleneck of online-order fulfillment,” Xin says. Having workers grab items for customers at stores may also be less efficient than warehouse fulfillment, he says, particularly because warehouses tend to have equipment and robots to help select items. Some retailers will invest more in store infrastructure and upgrade stores to miniwarehouses, as Walmart has done.

The continued rise of robots is another trend with which logistics professionals are wrestling. In the US alone, according to the Association for Advancing Automation, factories ordered nearly 30,000 robots during the first three quarters of 2021, a 37 percent year-over-year increase. Research also suggests that robots can make factories more efficient, by grabbing products from shelves and handing them to human workers to package. Could they help make factories more flexible, and thus chains more resilient?

The push to make supply chains greener adds yet another dimension. (See “The business case for sustainable supply chains.”) And figuring out how all supply-chain goals fit together will occupy companies for years, even when the pandemic becomes, hopefully, a distant memory. This will mean companies must be willing to take the steps to consider some of these ideas. They’ll decide whether they want all their suppliers in farflung locations or closer to home; they’ll weigh expenses and risks on the basis of their industry, size, and other factors. “There are existing frameworks for structuring supply chains and managing risk,” says DeHoratius. “The question is: Are companies ready to execute on those frameworks?”

More from Chicago Booth Review

More from Chicago Booth

Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.