July 2026
One trillion dollars. That’s what global data center capital expenditure is on track to exceed in 2026. JLL’s 2026 Global Data Center Market Outlook suggests total data center expenditures over the next five years could approach $3 trillion when tenant GPU and networking fit-outs are included. It seems unfathomable, yet the industry certainly isn’t starving for investment.
How can organizations get a handle on such enormous capital expenditures? A couple of differentiators stand out. One is having clarity around where and how to deploy that investment strategically; the other is maintaining visibility and control over cost and schedule for dynamic data center projects. Capital is showing up; the harder work is turning it into powered, cooled, permitted, operational capacity without losing control of the project.
Why the Investment Shift Is Happening
What’s driving this spike in digital infrastructure investment?
Much of it is coming from the evolving ways we’re all interacting with AI. In a few short years, we’ve gone from sifting through pages of blue-linked search engine results in response to a keyphrase query all the way to asking questions on AI platforms and getting fully detailed responses with supporting source links.
This shift in how we—as individuals and as organizations—use data is a sign of where the data center market is going. The compute power required to support that kind of interaction is far greater than what traditional data workloads demanded.
The numbers alone tell the story. AFCOM’s 2026 State of the Data Center report found average rack density reached 27 kW per rack, up from 16 kW last year—a 69% year-over-year jump and nearly four times the 7 kW reported in 2021. That kind of acceleration does more than just change what goes inside these facilities. It fundamentally changes how they have to be designed and built.
The infrastructure required to support that kind of AI-driven interaction is more power-intensive and intricate than what came before it. The U.S. Department of Energy and Lawrence Berkeley National Laboratory estimate that data centers consumed about 4.4% of total U.S. electricity in 2023 and could reach 6.7% to 12% by 2028; Goldman Sachs Research forecasts global data center power demand could rise 165% by 2030 compared with 2023 levels. Development is already moving beyond crowded primary markets and into secondary and tertiary markets that were barely blips on the radar a few years ago.
Those markets come with their own complexity—less established infrastructure, fewer experienced subcontractors, different utility interconnection paths and different permitting requirements. Getting ahead of that requires more strategic planning from the start. How organizations respond to that complexity, and how well they manage the capital behind it, is where the real work begins.
How Organizations Can Respond
With rising digital infrastructure investment and the rapid pace of change in specifications come higher stakes in the outcomes, as well as how projects are managed. AFCOM’s 2026 survey also found average facility size approaching 38 MW, up from 32 MW last year, while 74% of respondents plan to deploy AI-capable infrastructure and 72% expect AI workloads to significantly increase capacity requirements.
For construction executives, that means asking harder questions earlier. Does this market support the scale and timeline this project demands? Are current specifications designed to accommodate where rack density, cooling, and power delivery requirements are heading, not just where they are today? What happens to the budget and schedule if those specifications change mid-build? And perhaps most critically, what project controls are in place to manage a build of this value and complexity?
After all, data center specifications aren’t static. Take design, for instance. That stage alone reflects how dramatically requirements have shifted, as we’re seeing with rapidly advancing AI rack density and cooling systems that continue to evolve faster than most project timelines allow. The same AFCOM survey found 39% of respondents say their current cooling solutions do not meet all operational requirements; 36% have deployed liquid cooling, and another 28% plan to adopt it within the next 12 to 24 months.
At the capital levels flowing through data center construction right now, the consequences of accommodating a necessary change or catching a problem late are significant. Minor misses stop being minor when they collide with megawatt-scale schedules. What might have been manageable on a smaller project can be extraordinarily costly here. Having a clear, current picture of where a project stands—and the processes to act on that information quickly—makes a meaningful difference in outcomes.
That’s precisely where strategic clarity and rigorous cost and schedule control matter most.
The Bottom Line
A trillion dollars in capital is only as effective as the strategy and controls behind it. Deep pockets help, but they do not create megawatts, permits or certainty on their own. The data center market is moving fast, and it’s getting more complex, more expensive and more consequential with every passing year. The organizations that will navigate it well aren’t necessarily the ones with the deepest pockets. They’re the ones with the clearest picture of where to invest and the controls in place to manage it once they do.
How well are your project controls suited to evolve along with the growing data center market? If the questions raised here are ones your team is wrestling with, it may be worth taking a closer look at your current capabilities. Learn more or request a demo today.
———————————
About the authors
Bill Kleyman
Co-founder, CEO, Chairman of the Board, Apolo
Award-winning data center trailblazer and top cloud influencer. Bill Kleyman brings 20+ years of experience as a visionary in cloud, data centers, and AI, holding pivotal senior executive roles at Switch Data Centers, EPAM Systems, and MTM Technologies.
Andy Verone
Chief Strategy Officer, Contruent
With over 35 years of industry experience, Andy joins Contruent with a strong track record of executing innovative strategies, leading cross-functional teams, and transforming businesses. With a deep understanding of the customer experience, Andy has worked to accelerate technology adoption and technological growth in every role. Prior to joining Contruent as Chief Strategy Officer, Andy served as Global Vice President at Oracle, overseeing M&A activities, shaping product strategies, and co-founding Oracle’s Vertical Industry labs. His industry knowledge and breadth of experience across teams will be a critical asset as Contruent grows, innovates, and transforms projects for clients globally.