Historically, site selection was driven by proximity to customers, land availability, connectivity, and planning. Site selection now also involves data sovereignty, power, water and other infrastructure, environmental constraints, community acceptance, supply chain resilience and access to capital. Governments take a more active role in shaping digital infrastructure, and capital is entering earlier in the development lifecycle.
The result is a new location equation. Success now requires aligning infrastructure, regulation, customer demand, planning and environmental deliverability, social licence and capital. The most attractive sites are those where these factors are assembled into a deliverable, scalable development package.
A new site selection framework
Site selection starts with the intended use. Edge and latency-sensitive applications still prioritise proximity to users, fibre networks and major exchanges. AI training clusters and sovereign compute platforms can tolerate greater distance from users in exchange for the infrastructure needed for high-density computing.
As AI becomes a strategic national priority, regulatory frameworks are becoming a key differentiator between markets, and planning now sits at the heart of location strategy. Some governments are accelerating deployment in selected areas through planning support, infrastructure funding and access to land. Others are tightening controls in response to environmental concerns, resource constraints and community opposition.
In some markets, accelerated deployment quickly gives way to tighter controls, driven largely by community opposition.
Power remains the defining constraint
Power has emerged as the dominant constraint in many markets, followed by water supply and social opposition. Power is the gateway through which all other site-selection decisions are filtered: a location may satisfy customer, connectivity and commercial requirements but remain unviable if power cannot be delivered timely.
Investors globally prioritise locations where power infrastructure and planning strategy are aligned.
Governments are actively shaping where and how digital infrastructure is deployed.
In Asia-Pacific, governments are linking data center development to wider economic and sustainability objectives. Singapore uses structured programmes to allocate infrastructure capacity. Japan is deploying incentives to encourage development beyond Tokyo and Osaka. India has designated data centres as essential infrastructure, with states competing for investment through land and other incentives. In Australia, the focus on which projects can actually energise has driven regulatory reform, clean energy mandates, queue management initiatives and greater scrutiny of speculative projects.
In the United States, demand continues to outpace supply, pushing development beyond established clusters. Pressure on power, infrastructure and local communities is driving attention towards regions that offered greater capacity and planning certainty. But as demand escalates, regions are introducing measures to curb speculative projects, impose “bring your own generation” and emergency curtailment requirements, and audit water and land use by data centres, balancing rapid load growth against grid reliability and scarce natural resources.
Europe combines strong demand with increasingly sophisticated sustainability and reporting requirements, while development remains shaped by local planning controls, grid availability and environmental requirements. Spain’s draft rules illustrate the direction of travel: grid access for new data centres of 1 MW or more would be tied to resilience and digital sovereignty, energy and water efficiency, renewable power and disclosure requirements, with non-compliance risking loss of access permits.
The UK has designated data infrastructure as Critical National Infrastructure, emphasizing their importance to economic growth and digitalisation. The National Planning Policy Framework (August 2026) strengthens support for data centres and their locational requirements, including electricity network infrastructure. Newly designated AI Growth Zones offer selected sites streamlined planning and other incentives. Harder-edged legal reforms are speeding up consenting: data centres need not follow any prescribed consenting route, and new flexibility means planning strategy can bend to accommodate power strategy. Judicial review reforms aim to limit the ability of legal challenges to stall important projects.
The Middle East is increasingly encouraging data center development as part of broader digital economy and AI strategies. The UAE has risen to regional prominence by offering a stable environment for investment, significant tax incentives, and streamlined licensing procedures, underpinned by strong sovereign backing and abundant energy. However, data centres in the region must now confront regional conflict, as gulf-based data centres have been targeted as perceived extensions of US-aligned infrastructure.
Site acquisition structures are changing
Growing complexity is changing how sites are acquired, funded and developed. Few organisations can independently identify, secure, power, finance and operate large-scale developments, so collaborative models are increasingly important. Commercial models are evolving:
Joint ventures and strategic partnerships combine land, development capability, power access, supply chain buying power, operating expertise and institutional capital, allowing participants to share risk and accelerate delivery. Careful thought must be given to governance, deadlock resolution and how different contributions are valued and protected over the life of the venture, particularly where one party's contribution is front-loaded while another's accrues over time.
Vendor financing arrangements bridge the gap between land acquisition and project maturity. Deferred consideration, milestone payments, overage mechanisms and seller loans can reduce upfront capital requirements while aligning seller and developer interests.
Leasing structures are adapting to the evolving data center market. Build-to-suit arrangements, powered shell deliveries and phased energisation schedules have become standard in many regions as tenants seek to derisk deployment timelines while landlords manage infrastructure commitments. Sale-and-leaseback structures let operators recycle capital from stabilised assets into land acquisition, power procurement and AI-related infrastructure while retaining operational control through long-term leases. Data sovereignty considerations must be taken into account when selecting structures.
Forward funding and forward commitment structures bring institutional capital in much earlier than traditional real estate models, with investors funding development against agreed milestones or committing in advance to acquire completed assets. They parallel build-to-rent and purpose-built student accommodation, where investors fund development against pre-agreed specifications and milestones. Added complexity lies in defining completion and handover standards for highly technical facilities, including Tier III or IV certification, power usage effectiveness (PUE) thresholds and customer-ready fit-out requirements.
Equipment manufacturer financing and lease backstops are an increasingly important feature in some regions. Critical equipment manufacturers now act as financing parties for personal property, offering lease financing, deferred payment and equipment-as-a-service models that reduce upfront capital requirements. Some also provide credit backstops for tenant lease obligations, guaranteeing a portion of rent or providing a standby letter of credit, so that operators with thinner balance sheets can secure premises from institutional landlords. Structuring these backstops requires careful attention to the interplay between the primary lease, the backstop agreement and any equipment financing, including subordination, cure rights and remedies on tenant default.
Successful site acquisition is no longer simply about securing land, and the line between site selection and transaction structuring is increasingly blurred. Winning sites will be those where capital, infrastructure, community benefit and expertise can be brought together most effectively.
Read Ashurst Perkins Coie's extended thoughts on this topic here.
Supporting authors: Camarin Madigan, Alex Peace, Ismael Fernandez Anton, Alex Peace, Yi Ming Choo, Michael Condlyffe, Maria Antonia de Prada, Laura Morton, Neva N. Wagner, Daisy Choi, Kit Lee, Matt Taylor, Jane Hall.
Related stories
-
Sponsored Data centers and the social licence to build
Community and social impact management is not simply a risk to be mitigated, but an opportunity to build the durable trust and maturity the sector will need
-
Sponsored The new data center playbook
Why a sophisticated approach to collaboration, governance, compliance, dispute resolution and risk allocation is key to navigating an ever-evolving data center landscape
-
Sponsored The faster we listen, the faster we can move
“The data center industry doesn't have a permitting problem or a communications problem. It has a trust problem”
Comments