Look around and you’ll spot them here and there. Companies doing things that look like business-as-usual moves that support global growth but are actually something else: strategic steps to hedge against future threats to their value networks. These companies are onto something, and other businesses need to take note.
Consider: About a year ago, in February 2023, German car manufacturer BMW announced an investment of up to €800 million in a battery factory in Bavaria, after making similar investments in assembly in Debrecen in Hungary, San Luis Potosí in Mexico, and South Carolina in the United States. And in November 2023, Foxconn, an Apple iPhone manufacturing partner, announced plans to invest US$1.5 billion in operations to strengthen its workforce in India. Meanwhile, North American chip manufacturer Intel has plans to set up two new factories in Magdeburg, Germany.
At first glance, all these activities could be examples of tactical moves to broaden markets, or drive innovation, or lower costs. But BMW is combining its pivot to electromobility with regional access to the most critical technology. Apple is shifting production away from China in response to geopolitical and (yes, still) COVID-related disruptions. And Intel is responding to geopolitically driven regulations in the chip industry while tapping into pools of substantial planned government subsidies.
Globalization isn’t in retreat, but the forces affecting it have evolved (see HBR’s “The State of Globalization in 2023,” and these companies mentioned above are taking action to adapt. Specifically, they’re keenly aware of changing trade restrictions, data regulations, and industrial policy, and they’re weaving that awareness into their more traditional expansion goals, such as economies of scale and labour arbitrage. They know that local units must be able to adapt to any and all of those forces and that to be sure that’s possible, they need to have flexibility throughout their value chains. They know that if their suppliers are geographically concentrated, they need to be ready with solid supplier relationships in new areas. And if their market is similarly concentrated? They want to be sure that their customer base is broadening by the day.
Ultimately, they’re developing more clustered, decentralized value chains to meet regional policy requirements even as they’re taking greater control over the big picture, so that they can leverage various supply locations as needed and still be compliant in the markets they’re serving. At these and other forward-thinking companies, top-down policies are starting to guide increasingly complex regional implementation.
Most executives know that this sort of approach is needed to thrive. According to research undertaken by Accenture Research, 78 per cent of the over 2,300 C-suite executives surveyed globally expect increasing fragmentation. Yet as our follow-up interviews indicated, many feel unprepared to handle it. As one C-suite executive from a capital equipment manufacturer told us: “We manufacture in China and have a strong supply base. In the near term, through 2030, we’ll be there. But geopolitical conditions will require us to diversify at some point… We’re just one small controversy away from catastrophe.”
Learning from Schneider Electric
Schneider Electric SE—the leader in the digital transformation of energy management and automation, headquartered in France and with businesses in Africa, Europe, the Asia-Pacific, the Middle East, and North and South America—offers an example of this new strategic positioning in more detail. The company’s multi-hub “glocal” (global-and-local) set-up is striking an effective balance between different requirements and customer preferences, and the need to help ensure resilience, growth, and cost effectiveness.
“We are the most local of global companies. Our multi-hub approach effectively balances global and local needs, ensuring resilience, growth, and scalability while boosting sustainability,” says Peter Weckesser, chief digital officer, Schneider Electric. (Disclosure: Schneider Electric is an Accenture client.)
Specifically, four hubs—one each in North America, Europe, India, and China—serve local customers with the support of local suppliers. These hubs adopt Schneider Electric’s products and offerings to the specificity and standards of local markets (e.g., NEMA in North America, IEC in Europe, and CCC in Asia-Pacific), regulations (e.g., data and cybersecurity), operating conditions, and design specificities. (That’s important: According to the latest available data, 137 countries out of 194 researched have legislation in place to regulate data transfer, data processing, and the protection of data privacy.)
Additionally, these hubs utilize global R&D platforms and architectures. By operating this way, if parts of the supply chain depend on components subject to market restrictions or regulations, Schneider Electric can meet that challenge while keeping other parts of the supply chain—and processes overall—running.
Schneider Electric uses a single operating model, with global business units and functions (marketing, digital, finance, HR, supply chain, and bespoke software) and with operations close to the customers it serves. Unifying bespoke software presents both a cost-saving and performance-improving opportunity but also a risk, given the need to separate regions for regulatory or other purposes. In this way, the company maintains a consistent brand value proposition, processes, and product standards across the business, bringing simplicity and cost efficiencies to each region.
An added benefit is the boost to sustainability. In 2021, 92 per cent of the company’s supplied goods came from the same region as its manufacturing sites, and 80 per cent of its sales were produced in the same region as its customers, cutting down on transportation expenses and energy consumption.
The tech that enables the strategy
What helped Schneider Electric create this sort of structure? A progressive approach to enterprise-wide technology adoption. The company’s commitment to building and maintaining a strong digital core enables it to mobilize technologies such as digital twins that make its central operating model and hub structure run smoothly. That core is also making it easier for Schneider Electric to experiment confidently with emerging technologies such as generative AI.
Paradoxically, when infrastructure technology (especially information technology that was once specific to an operating unit or country) is more unified across a multinational corporation, that very unification allows companies to more quickly adjust operations to match local requirements. They can expand and contract to create operational “islands” as needed.
But they can also lean on that unification in other ways. One of the largest areas of regulatory impact we’re seeing on the technology stack across the companies we studied and work with is access to the underlying chip technology needed to productively use data at scale.
Increasingly, for example, governments are developing controls to restrict the export of the most advanced technologies that power AI. In the near term, these kinds of restrictions will affect where the most advanced AI technologies can be deployed. And that will wreak havoc on companies with siloed information technology structures as they rush to understand how to deploy generative AI across their internal activities and in their products and services. One challenge with large language models (LLMs) is that the cost to run per query is much higher because of the large number of calculations required. Another challenge is that firms will need to control access to their proprietary data, which means they will need new infrastructure that can train LLMs and other generative AI systems internally to avoid data leaks like the ones that have already occurred.
Or take cloud computing, which creates exposure to regulations regarding data residency such that certain data must not be copied outside of regions such as the European Union. More directly, technology regulation is affecting what wireless communications technology (5G and soon 6G) can be deployed, with Australia, Japan, and the United States all restricting the import of technology from Chinese firms such as Huawei.
Interestingly (though this nugget is a bit of a tangent), these restrictions on data access and residency are driving innovation at the algorithmic level. For example, long-held restrictions on data use to protect privacy in health care have led to the development of systems to train AI models across dispersed data sets.
Getting started
If your company isn’t ahead of the game in staging to protect its value chain, it’s certainly not too late to begin, but it is time to move with urgency.
Start by thinking differently about what a global footprint means. Study the policy landscape with the same intensity with which you study market attractiveness and consumer preferences. Think about what kind of value you can get from where, while safeguarding your global operating model. Consider the following three activities:
First, map where your company operates with just a few data clusters; also map the areas where it holds data in many disparate places. Ask: What is the extent to which having more data from across the organization has (or could have) a bearing on outcomes in those areas?
For example, consider applications that track utilization, meetings, and meeting-room reservation schedules. The value an organization derives from the data is not likely to increase significantly from access to more data, so it may not matter very much if your company stores such data locally or within country jurisdictions. In contrast, data used for industrial or service process automation is probably collected from many locations across the company and pooling the data globally might yield process improvements that would not be possible when using smaller data sets. However, regulations might require that such data be stored and utilized within jurisdictions that prevent pooling. Companies facing this data use case need to be aware of the potential for regulatory changes and should institute a way to monitor their evolution so that restrictions won’t have a major impact on performance, potentially through pooling certain types of operations within jurisdictions, or, more feasibly, by investing in systems that can learn from data without moving it across jurisdictions.
And for digital twins or factories with long supply chains, where data resides in many separate jurisdictions? Aggregating data from across the company can create enormous value, and sudden restrictions could have significant negative consequences. Therefore, these areas require active and continuous regulatory monitoring and mitigation planning, should regulations change.
The second activity is to ask how you can capture value from your company’s data assets in inventive ways. Most firms begin with operational or expense improvements within their organizations, and that is a good way to justify early investments. Firms might also think about selling data itself. We would proceed with caution in this case because we’re more interested in how companies can use data for other purposes—say, to create a new product or service. For example, firms might invest in a rigorous mapping of their carbon footprint (and the environmental impact of their products in use) and find low-cost ways to reduce their footprint. Given that other firms face similar challenges, this might become a monetizable service offering.
Third and finally, consider how to leverage government investments in frontier technologies. No longer can companies make decisions from a regulatory point of view alone about where to locate manufacturing facilities, how to design supply chains, and where to store and process data. Today, they must also consider what national support, such as subsidies, is available; how critical that support is to their business; and how leveraging that support will affect their strategic plans.
Not long ago, standardization, industrialization, and scalability were the hallmarks of a smart approach to globalization. Today, in a more multi-layered world, companies need to be able to take a multi-layered approach to thrive.
