Rethinking data centre site selection

The roundtable at the Data Centre Leaders Summit felt different to many industry conversations. Not because of a technical breakthrough or policy change, but because of the nature of the discussion. It was grounded, sometimes uncomfortable and shaped by people actively delivering data centre projects across Australia.

Developers, planners, architects, government representatives and infrastructure providers brought different perspectives, yet a common theme emerged: Data centres are becoming part of communities, whether planned that way or not. This places greater weight on decisions made early, particularly during site selection, where planning risk, delivery certainty and community outcomes begin to take shape.

This shift reframed the conversation around a critical question: How can site selection deliver value beyond the facility itself?

Participants broadly agreed that data centres can support digital economies, create employment opportunities and contribute to energy transition initiatives.

This creates a disconnect. Public narratives tend to focus on power use, scale and perceived impacts rather than potential benefits. As a result, conversations tend to be reactive, occurring once concerns emerge rather than earlier in the planning process.

Data centre conferences, and particularly focused round table discussions, provide valuable space for dialogue, but their reach is limited. If trust is the goal, the data centre industry needs to develop a robust community engagement strategy to enable greater transparency about intent, constraints and potential community benefits to help shift conversations from concern to collaboration. For developers and operators, this translates into fewer planning challenges, stronger community acceptance and more predictable project timelines, Early engagement also helps move stakeholders closer to solutions, creating opportunities to work together to address concerns and deliver shared outcomes.

Traditionally, site selection has been driven by factors such as latency, access to power, planning pathways and cost. These considerations remain critical, but they are no longer sufficient on their own.

Despite advances in technology, proximity to major cities continues to influence many location decisions. Regional locations may offer lower operating costs, greater access to land, development flexibility and economic benefits for regional communities, yet concerns about performance and connectivity still draw projects toward metropolitan fringes.

Participants questioned whether this approach needs to evolve. If data centres are to move beyond Sydney and Melbourne, infrastructure planning, local government knowledge and state government policy alignment will need to evolve in parallel. Without this alignment, developers remain constrained and communities miss potential investment.

These pressures reflect a broader shift in how data centres are planned and delivered, where site selection is increasingly linked to infrastructure capacity, operational requirements and long-term adaptability. We explore this system view further in Infrastructure solutions for an AI-driven world.

One of the most candid parts of the various conversations focused on the relationship between government, infrastructure providers and developers. Government representatives called for earlier and more open engagement on future plans, while infrastructure providers highlighted the challenge of planning upgrades without visibility of future demand. Developers, in turn, pointed to commercial sensitivity and competitive pressures.

This misalignment has practical implications. Delays in infrastructure provision can slow projects. Uncertainty can heighten community concern. Opportunities for shared outcomes are missed when conversations begin too late. Better outcomes rely on new ways of working that balance commercial realities with broader planning objectives. Earlier, high-level conversations, even when details remain commercially sensitive, can help align infrastructure investment with anticipated growth, reduce delivery risk for project teams and support more coherent long-term planning for government and communities.  

Community-centred site selection was a recurring theme, the conference roundtables also explored what this means in practice.

At its core, it is about recognising that data centres increasingly sit within communities rather than apart from them. As facilities become more distributed and, in some cases, smaller, their interface with surrounding neighbourhoods becomes more direct. This brings responsibility as well as opportunity.

Waste heat was a consistent topic. The idea of capturing and reusing waste heat for nearby community facilities is technically viable and socially beneficial. However, opportunities are often stalled by unclear ownership, complex approvals or ideas emerging too late in the process. The takeaway was not that these outcomes lack value, but that clearer frameworks are needed to make them achievable. Policy that provides clear pathways for councils and infrastructure providers can help enable earlier collaboration and reduce uncertainty.

Considering opportunities such as shared infrastructure or energy reuse during site selection makes them easier to integrate and more likely to translate into viable outcomes.

The pace at which data centres are being delivered also shaped the conversation. Demand is high and timelines are tight. Several participants noted that pressure to build quickly can limit the ability to consider broader community contributions.

When delivery speed becomes the primary driver, minimum compliance can become the default. Broader benefits are harder to prioritise, not because they lack value but because the system does not always allow space for them. Infrastructure constraints add another layer of complexity. Power, water and transport networks are often struggling to keep pace. Developers face delays, infrastructure providers struggle with uncertainty and communities experience disruption without always seeing longer-term benefits.

The discussion suggested that slowing projects is not realistic, but thinking earlier is. When community outcomes are considered from the outset, they shape site selection and design rather than becoming optional additions later. 

Toward the end of the conversation, participants raised a longer-term question: As technology evolves, what happens to the infrastructure we build today?

Future facilities may require less space or operate more efficiently, potentially leaving oversized assets or under-used infrastructure. This is not a reason to delay investment, but it highlights the need for whole-of-system thinking.

Data centres sit within a broader ecosystem, and site selection decisions made today will shape communities for decades. Considering adaptability, reuse and integration during early planning is just as important as meeting current demand, helping projects respond to changing technology and community needs over time.

A clear takeaway from the roundtable was the need to move beyond industry closed conversations. The sector is delivering more than it often receives credit for, yet these contributions are not always visible. At the same time, genuine gaps remain.

Community-centred site selection is not about choosing between development and community outcomes. The two are linked. When data centres are planned with communities in mind, they can support local economies, contribute to shared infrastructure and build trust.

The conversations at the Data Centre Leaders Summit reflected a willingness to engage with complexity. The next step is to carry these discussions wider, bring more voices into the room and translate intent into action. Facilitators, designers, developers and advisors all have a role to play in shaping how this sector evolves. Not only by delivering projects efficiently, but by helping data centres become valued neighbours withing the communities they serve.