The foundations of early 21st century manufacturing network strategy are starting to look shaky. Economic, technological, and political shifts have undermined the traditional advantages of globalized supply chains, just as the definition of supply chain value is evolving. Delivering products at the right time, the right quality, and the best cost is no longer enough. Organizations now need networks with the resilience to tolerate shocks and the agility to respond to demanding customers and fast-changing markets. They also want their supply chains to support their efforts to achieve environmental and social sustainability.
Over the past decade, automation and the rising cost of energy and raw materials mean that factor-cost advantages are becoming less significant than they were. Add in the effect of new tariffs, and transport costs that have more than tripled on some major ocean routes, and the cost advantage of global production shrinks still further.
Companies are becoming more concerned about the disadvantages of globalized networks too. Long supply chains are more vulnerable to disruptions, which are already costing the average company the equivalent of 42 percent of one year’s profits every decade. Thin supply chains, where critical inputs are sourced from a single supplier and substitutions are difficult, tend to exacerbate the impact of problems and extend recovery times. Time zone differences, language barriers, and long transit times make supplier collaboration in product development and quality assurance more challenging.
The same goes for customers. Global shipping can add a month or more to delivery times that are already frustratingly long. And consumers are increasingly sensitive to the “invisible” features of a product, such as its carbon footprint or the labor conditions in its supply chain.
These challenges don’t mean that wholesale reshoring is going to be the best solution for every supply chain, however. Tight labor markets in many regions are making it tricky for companies to expand their operations or set up new ones. In some sectors, economies of scale or skill make it sensible to concentrate production in large plants that serve global markets.
Manufacturing’s new network models
With so many elements in play, and so much uncertainty, companies need a better way to evaluate their current and potential future network strategies. One promising approach is with sophisticated, industry-specific models based upon an expanded definition of the value delivered by the supply chain.
These models attempt to quantify the impact of all the network-related factors that affect an organization’s performance, from changing input prices to supply disruptions or the cost of carbon emissions. They allow companies to examine critical network design trade-offs, such as regionalization versus globalization or on shore versus nearshore operations, and help them understand the likely impact of key uncertainties.
The four sections that follow—“Consumer electronics: Localization is key,” “Pharmaceuticals: Scale and efficiency matter most,” “Capital medical devices: Automation helps maintain current margins,” and “Apparel: Customers’ willingness to pay matters most”—show examples of this modeling approach applied to hypothetical companies in several different sectors, with each designed to be representative of its sector. The results can be startling. For the consumer electronics example, the model predicts that a business-as-usual approach to its network will see its profit margins disappear entirely by 2030, despite the near-term productivity improvements offered by automation.
But radical reconfiguration of the company’s supply footprint could reverse this erosion of its margins. Moving some or all its supply chain to the US may boost profitability directly, while also increasing the flexibility and responsiveness of the supply chain.
Of course, it is neither technically nor financially viable to make such a shift overnight. Localized supply chains require capable local suppliers. Building new production facilities takes time and capital. Upstream industries may face similar constraints. Energy-intensive businesses such as chemicals and basic materials may need time to secure reliable sources of low-carbon power and adapt their process to use it, for example. And regionalization won’t work for every business. For a hypothetical company manufacturing active pharmaceutical ingredients (APIs), a regional production strategy only decreases its margins, suggesting that productivity and yield improvement are the best ways to beat rising costs.
The real power of this sort of modeling lies in helping companies determine a long-term vision for their networks. If an organization recognizes that localization or the switch to low-carbon energy will be advantageous over the next decade, it can begin to work with existing suppliers to develop the necessary capabilities. And when it needs to make new asset investments or identify new sources of supply, it can make those decisions with its long-term goals in mind.
The granular picture of cost evolution provided by the model can also help organizations to target efforts for improving the performance of their existing networks, for example by accelerating the introduction of automation or optimizing product flows to reduce logistics costs and emissions.
Modeling the evolution of supply networks in this way also demonstrates the importance of a broad, cross-functional approach to manufacturing optimization. Changes to materials or product designs can have a big impact on labor productivity and carbon emissions, for example. And those choices must be made alongside decisions about where and how production is done.
Consumer electronics: Localization is key
Exhibit 1 shows the output of a value-chain modeling exercise for a medium-size consumer electronics player. The company’s manufacturing and supply footprint is currently concentrated in Asia, but its customers are in Asia and North America. The model assumes that demand, factor costs, and duties all evolve in line with their recent five-year trend until 2030. Logistics costs are assumed to remain at today’s levels, and the model assumes that competitive factors prevent the organization from increasing its sales prices.
Under this scenario, higher duties and factor costs will cut the organization’s profit margins by seven percentage points over the next eight years. Add the impact of supply chain disruption (at cross-industry average levels) and introduce carbon pricing, and margins drop by four additional percentage points.
If the organization retains its current supply network, the primary lever available to control rising costs is productivity improvement through automation. That may not be enough. Even aggressive reliance on automation across the entire value chain would not be sufficient to maintain profitability once supply chain disruption and carbon pricing are considered.
For this company, manufacturing network changes offer a much more compelling option. If, alongside greater automation, the organization were to transfer the whole of its supply footprint to North America, margins would be expected to increase by 20 percentage points.
Such a wholesale transfer is unlikely to be possible in the short term since North America does not currently have a significant supply base to produce high-volume, low-cost electronic components. A partial nearshoring approach is more viable, with assembly of mechanical and electronic subsystems transferred to facilities in North America. The model suggests that doing this would increase margins by four percentage points, allowing the organization to maintain its margins at close to today’s levels (Exhibit 2).
Pharmaceuticals API: Scale and efficiency matter most
Exhibit 3 shows the output of a value-chain modeling exercise for a pharmaceutical company. This company’s API manufacturing and supply footprint is currently concentrated in Asia. It manufactures finished drug products using regional networks serving customers in Europe and Asia. The model assumes that demand, factor costs, and duties all evolve in line with their recent five-year trend until 2030. Logistics costs are assumed to remain at today’s levels, and the model assumes that competitive factors prevent the organization from increasing its sales prices.
Under this scenario, the model suggests that the margins on the company’s prescription (Rx) products are likely to fall by two percentage points over the next decade. Margins on generic (Gx) products are expected to fall slightly further, with a drop of up to seven percentage points.
In this company’s case, shifting API production closer to end customers does little to address the problem. That’s because potential savings in transport, customs duties, and carbon emissions are more than offset by the higher cost of operating multiple small API plants. Industry benchmarks suggest that API conversion costs at a small plant are up to 40 percent higher than at a medium one. As a result, the current API production footprint is five percentage points more profitable than a regionalized approach.
Shifting that balance would require several simultaneous changes. First, demand for the company’s Rx products would need to grow sufficiently to require regional API plants of above median size. Industry benchmarks also show that economies of scale trail off rapidly as plants get bigger, so two medium plants are almost as cost-effective as one very large one. Second, productivity would need to increase by 35 percent to be cost-competitive—an aggressive ambition that might be possible through digital-yield optimization, improved batch processing, and related investments. Third, the company’s product portfolio would need to be especially sensitive to disruption risk.
Under the regionalized scenarios, input costs are still five times larger than transport costs. Producing inside the European Union would reduce margins by seven percentage points compared with lower-cost neighbors, for example. Customs duties would need to increase by (a likely unrealistic) 40 percentage points to close the gap.
The case for a North American regional network is similarly difficult to make. Producing APIs in the US Midwest would reduce margins by 18 percentage points. Shifting American production to South or Latin America would be less disadvantageous, thanks to lower relative input costs in these regions. The resulting gap of five percentage points might be easier to close through growth and efficiency improvement measures.
In the absence of overwhelming concerns about supply chain risk or the need to dramatically improve responsiveness, this pharma company should probably stick to its global production strategy, while investing in automation and productivity improvement measures across its sites to offset rising costs.
Capital medical devices: Automation helps maintain current margins
Exhibit 4 shows the output of a value-chain modeling exercise for a medical-device player producing capital medical equipment. The company’s assembly and test footprint is concentrated in the United States, with supply footprint split between Asia for electrical components and North America for mechanical components. Customers are predominantly in the United States. The model assumes that demand, factor costs, and duties all evolve in line with their recent five-year trend until 2030. Logistics costs are assumed to remain at today’s levels, and the model assumes that competitive factors prevent the organization from increasing its sales prices.
Under this scenario, higher duties and factor costs will cut the organization’s profit margins by nine percentage points over the next eight years, if no changes to the current manufacturing network footprint are made. Add in the impact of supply chain disruption (at cross-industry average levels) and introduce carbon pricing, and margins drop by 14 percentage points.
If the organization retains its current supply network, using automation is sufficient to maintain current margin levels. For this company, further regionalization in the supply chain in North America (such as locating suppliers closer to assembly plants) does not significantly improve long-term performance, as increased resilience and reduced transport costs are offset by additional overhead from duplicated assets (Exhibit 5). However, nearshoring could be investigated further, given its potential to allow shorter lead times and derisk the supply chain, preventing disruptions that could directly impact patients. Nearshoring assembly operations to certain locations in North America, meanwhile, could be an option to improve margins in an inflationary environment.
Apparel: Customers’ willingness to pay matters most
In Exhibit 6, an apparel player’s raw-fiber processing is concentrated in East Asia, while its fabric and cut-and-sew footprint is in South Asia. Customers are in Europe and Asia. The model assumes that demand, factor costs, and duties all evolve in line with their recent five-year trend until 2030. Logistics costs are assumed to remain at today’s levels, and the model assumes that competitive factors prevent the organization from increasing its sales prices.
Under this scenario, the model suggests that if no changes to the current manufacturing network footprint are made, margins are likely to decrease by nine percentage points over the next decade—even if the manufacturer deploys a high level of automation. The drop is mainly driven by higher costs related to duties, utilities, and carbon. Accordingly, additional levers need to be pulled in the form of increased efficiency, agility, and perceived product quality.
Transitioning from a global to regional manufacturing network would not by itself increase margins by much: they would still likely fall by nine percentage points because of factor-cost increases, including utilities. However, nearshoring could reduce lead time to the European market by about 20 days; higher costs could potentially be recovered, at least in part, by reducing overstock, markdowns, and stockouts. For a high-margin product, such as athletic T-shirts made from polyester, customers’ higher willingness to pay for perceived quality or better environmental, social, and governance impact may suggest that nearshoring could be a viable strategy.
