Data centers have quietly become one of the most strategically important asset classes in commercial real estate. They sit at the intersection of technology, energy infrastructure, capital markets, and land use planning, requiring a very different mindset than traditional industrial, office, or residential development. In this episode of The Real Finds Podcast, Gordon Lamphere sits down with Anita Verma-Lallian to unpack what data center investing actually looks like in practice, why Arizona has emerged as a critical market, and how developers are navigating power, capital, and exit strategy in a low-liquidity environment.
Anita’s perspective is grounded in long-term development experience. Coming from a family with more than 40 years in land development, she has seen multiple real estate cycles and asset classes evolve. What began with residential and industrial real estate development has now expanded into large-scale data center campuses, film studios, and infrastructure-heavy projects that demand both patience and precision.
Why Data Centers?
Data centers are not an obvious pivot for most developers. They are capital-intensive, technically complex, and dependent on infrastructure that cannot be easily replicated. Anita explains that her firm’s entry into data centers began organically. Arizona was already attracting hyperscalers and large-scale users, and the firm was developing industrial land with infrastructure requirements that closely mirrored what data centers need.
That overlap meant large sites, heavy utility coordination, and long planning horizons, making the transition a logical choice. What accelerated interest was the surge in demand driven by cloud computing and AI. Unlike many traditional asset classes, data centers have continued to attract capital even as broader commercial real estate markets have faced pricing pressure and reduced liquidity.
What Makes a Good Data Center Deal
At the core of every viable data center deal is power. Access to electricity, both in volume and reliability, is the gating factor. Anita emphasizes that proximity to transmission lines, substations, natural gas pipelines, or alternative energy sources often determines whether a site is feasible at all. In many markets, utilities are quoting timelines of 10 to 15 years for full grid power delivery, forcing developers to think creatively.
Land scale is the next critical requirement. The projects Anita is pursuing are measured in hundreds or even thousands of acres, designed to support multi-gigawatt campuses over time. These are not infill parcels. They require zoning support, political alignment, and enough separation from residential areas to avoid community opposition.
Pricing still matters, but unlike traditional industrial deals, it is often secondary to infrastructure feasibility. A cheap site without power is functionally worthless for a data center, while a more expensive site with energy access can justify its cost many times over.
Search, Not Shovel-Ready
One of the key takeaways from the conversation is that truly “shovel-ready” data center sites are rare. Most viable sites require significant repositioning. Anita shared an example of a 2,000-acre site originally entitled for residential development but lacking water infrastructure. Residential no longer made sense, but data center development did.
That kind of pivot is increasingly common. Developers are not just finding sites—they are creating them. This involves land assemblage, entitlement work, utility studies, and years of coordination with power providers. Data center development is less about speed and more about strategic patience.
Capital in a Low-Liquidity Market
While much of commercial real estate has struggled with limited liquidity, data centers have largely bucked that trend. Anita notes that capital, particularly from tech-focused investors, continues to seek exposure to data center development. These investors often bring deep technical knowledge of how data centers operate, while developers contribute land expertise, entitlement strategy, and execution.
This collaboration between tech and real estate capital is becoming a defining feature of the asset class. It also explains why data centers remain insulated from some of the interest rate sensitivity affecting multifamily or office development. The end users, hyperscalers, and AI-driven companies are less constrained by traditional financing metrics.
Exit Strategies: Build, Sell, or Hold
Exit strategy in data center development is not one-size-fits-all. Anita describes a flexible approach that varies by project. Some sites are assembled and sold to larger platforms that take over vertical development. Others are advanced further, entitled, powered, and partially built, before being sold or leased.
In certain cases, the strategy may involve long-term ownership and leasing, while in others, an earlier exit makes sense if pricing justifies it. The common thread is optionality. Successful data center developers preserve multiple paths to liquidity rather than committing to a single outcome too early.
Multi-Tenant vs. Single-Tenant Campuses
Given the scale of modern data center campuses, multi-tenant development is often the default. There are relatively few groups capable of absorbing an entire multi-gigawatt site in one transaction. Even major players like Microsoft or Google typically expand incrementally, adding capacity to existing footprints.
That said, demand can surprise. Anita notes that certain users, particularly AI-focused platforms, may require massive, consolidated sites. Developers must therefore design campuses that can accommodate both scenarios, preserving flexibility without overcommitting infrastructure prematurely.
Power, Water, and the Arizona Question
For Midwestern investors unfamiliar with Arizona, concerns about water often dominate the conversation. Anita pushes back on this assumption. Modern data centers are far more water-efficient than earlier generations, and water has not been a limiting factor in the projects her firm has pursued.
Power, however, remains the central challenge everywhere. Developers are increasingly exploring bridge solutions, including natural gas generation, solar, and proximity to nuclear facilities. In Arizona, political support for new energy infrastructure, including major natural gas pipelines, has expanded the universe of feasible sites.
These solutions are not universal, and cost remains a critical consideration. Developers must weigh sustainability, reliability, and economics on a site-by-site basis.
The Intersection of Data Centers, AI, and Film Studios
One of the more unique aspects of Anita’s platform is the crossover between data center development and film studio projects. At first glance, these may seem unrelated. In reality, both are infrastructure-driven, warehouse-scale developments with increasing reliance on data processing and AI.
Proximity to data centers can reduce latency, improve production efficiency, and support advanced workflows in media production. As AI becomes more embedded in editing, rendering, and post-production, this convergence is likely to accelerate.
Looking Ahead: AI and the Future of Real Estate
When asked what will change most in real estate over the next decade, Anita points to AI-driven site selection, infrastructure planning, and operational efficiency. Real estate will become smarter, not just in how buildings operate, but in how they are conceived, evaluated, and delivered.
For young professionals entering the industry, her advice is simple: take calculated risks and do not wait for perfect conditions. Real estate is cyclical, and opportunities rarely announce themselves clearly in real time.
Data center development embodies that lesson. It is complex, capital-heavy, and uncertain. But for those who understand the infrastructure, partnerships, and timing involved, it represents one of the most consequential frontiers in commercial real estate today. If you’re interested in learning more about how data center or data center adjacent investing can factor into your commercial real estate plans our commercial real estate agents would be happy to help!