For much of the past decade, companies were told that scale could be added almost on demand. Computing capacity could be rented, suppliers could be switched, transport could be booked, and capital could be raised when an opportunity appeared. That assumption is becoming less reliable. In a growing number of industries, the binding constraint is no longer demand. It is the physical and institutional infrastructure required to serve that demand.
The shift is visible across digital networks, electricity systems, logistics, industrial equipment and specialised labour. The OECD’s 2026 productivity indicators show that investment has increasingly tilted toward ICT and research and development, while overall investment rates remain below their pre-global-financial-crisis average. That combination matters: economies are asking more of a capital base that is being upgraded unevenly.
For businesses, this creates a different strategic problem. Growth plans must increasingly be tested against the availability of power, data-centre capacity, transport links, components, approvals and skilled people. A company may have customers ready to buy and financing ready to deploy, yet still be unable to expand on schedule.
The bottleneck has moved upstream
Traditional growth planning often starts with demand forecasts and works backward to production. Infrastructure constraints reverse that logic. Management teams increasingly need to ask what can physically be delivered before deciding how much demand to pursue.
This is particularly evident in digitally intensive businesses. The OECD’s work on digital transformation and business investment notes that digital adoption can change both the level and composition of investment. Artificial intelligence adds another layer because models may be software, but the systems supporting them require data centres, networking equipment, power and cooling.
The same principle applies outside technology. Manufacturing expansion can be limited by grid connections or specialist machinery. Distribution businesses can be limited by warehouse availability or transport corridors. Financial institutions can be limited by data architecture, cyber controls and regulatory approvals. In each case, the scarce input sits upstream of the customer transaction.
Power is becoming a business variable
Electricity used to be treated as a utility line item. For energy-intensive operations it is becoming a strategic constraint. The International Energy Agency’s Electricity 2026 report highlights continued growth in electricity demand, with data centres and other new sources of load increasing the importance of generation and network investment.
That does not mean every company needs to become an energy company. It does mean location decisions, expansion timetables and technology strategies increasingly depend on the reliability and cost of power. A site that appears attractive on rent and labour costs may be less valuable if grid access is delayed or if power quality cannot support critical equipment.
This changes capital allocation. Businesses may accept higher upfront costs for sites with dependable connections, invest in efficiency to release capacity, or sequence expansion around available power rather than around sales targets alone.
Connectivity is not the same as capacity
Global connectivity has expanded rapidly, but headline coverage can hide local constraints. The International Telecommunication Union’s Facts and Figures 2025 tracks continued growth in internet use while also documenting persistent gaps in access and quality. For companies, the issue is increasingly not whether a market is connected at all, but whether networks have the latency, resilience and redundancy required for modern operations.
The distinction matters for cloud services, remote operations, real-time payments, industrial automation and AI. A business process that works in a head office may fail when deployed across sites with weaker infrastructure. Resilience therefore depends on designing systems for the actual network environment rather than the theoretical one.
Logistics has become an operating moat
The pandemic made supply chains a board-level issue, but the longer-term lesson is broader. Efficient logistics infrastructure can determine how much working capital a business needs, how quickly it can respond to demand and how much inventory it must hold. The World Bank’s Logistics Performance Index provides a reminder that customs efficiency, infrastructure quality, shipment reliability and tracking capabilities vary significantly across markets.
When logistics are reliable, businesses can operate with tighter buffers. When they are not, the same level of customer service requires more inventory, more time and more cash. Infrastructure therefore affects not only delivery performance but also the balance sheet.
This creates an advantage for companies that understand their infrastructure dependencies in detail. A competitor may copy a product or pricing model, but it is harder to replicate a network of dependable suppliers, routes, facilities and contingency arrangements.
Capacity is becoming an option, not just a cost
Corporate finance traditionally treats unused capacity as inefficiency. That view is too simple when expansion lead times are long. Spare warehouse space, reserved computing capacity, secondary suppliers or underused production lines can function as options: they have a carrying cost, but they allow a business to react when demand or disruption arrives.
The value of that option rises when the future is uncertain and new capacity takes longer to secure. The World Economic Forum’s Global Risks Report 2026 underscores the range of economic, technological and environmental disruptions companies may need to absorb. The practical implication is not that firms should maximise redundancy. It is that resilience should be priced deliberately rather than dismissed as waste.
Growth strategy is becoming more physical
The most important change may be conceptual. Many corporate strategies of the past decade were built around intangible assets: brands, software, platforms, data and intellectual property. Those assets remain critical, but their value increasingly depends on physical systems beneath them.
An AI service depends on chips and power. E-commerce depends on warehouses and transport. Digital banking depends on telecoms and data centres. Advanced manufacturing depends on specialised equipment, energy and logistics. The more digital the customer experience becomes, the easier it can be to overlook the physical infrastructure that makes it possible.
Boards may therefore need to bring infrastructure mapping into strategic planning. The relevant questions are straightforward: Which inputs have the longest lead times? Which are concentrated among a small number of suppliers? Which can be reserved in advance? Which can be substituted? Which failures would stop revenue rather than merely raise costs?
What this means
The next phase of corporate growth may be defined less by who can generate demand and more by who can secure the capacity to fulfil it. That does not make growth slower by definition. It makes execution more dependent on preparation.
Companies that treat infrastructure as a strategic asset can make earlier decisions about location, suppliers, power, connectivity and redundancy. Those that continue to assume capacity will be available when needed may discover that the most expensive constraint is the one that appears after customers are already waiting.
References
• OECD Compendium of Productivity Indicators 2026 — Investment
• OECD — Business investment in the face of the digital transformation
• World Bank — Logistics Performance Index
• International Energy Agency — Electricity 2026
• International Telecommunication Union — Facts and Figures 2025