Summary
- Balfour Beatty does not plan to replicate its US data centre construction growth in the UK or continental Europe at current margins.
- Its US business secured around $350m of data centre work in the first half, with approximately $1bn more awarded during the year.
- The stance exposes a construction constraint behind Europe’s development pipeline: customer demand can remain strong while contractors reject risk that is not adequately priced.
Balfour Beatty is declining to chase UK and continental European data centre construction at current margins, even as the sector becomes a larger part of its US buildings order book.
The contractor secured around $350m of US data centre work during the first half of 2026 and said approximately $1bn of further awards had been secured during the year but had not yet fully entered the order book. US construction revenue rose 19% to £2.48bn over the period.
Chief executive Philip Hoare has said the UK and continental European data centre construction markets are too competitive at prevailing margins to justify a comparable expansion. The position separates strong underlying demand for new capacity from the returns available to the contractor taking responsibility for delivery.
A busy market can still be a poor contract
First half figures already showed data centres taking a larger share of Balfour Beatty’s US workload, including growth beyond its established north-western US operations into Virginia.
The group’s wider strategy places heavy emphasis on selective bidding and risk adjusted returns. Data centre construction fits that model only where contract terms, customer relationships, programme, and margin leave enough protection against the complexity of mission critical delivery.
A large data centre is unusually dependent on mechanical and electrical packages. Transformers, generators, switchgear, UPS systems, busway, cooling plant, pumps, controls, pipework, and specialist cabling account for a substantial share of project value, while commissioning has to prove that those systems operate together under normal and failure conditions.
One late equipment package can interrupt a programme even when the building shell is substantially complete. Design changes driven by customer density can also move through electrical distribution, cooling, controls, and structural requirements after procurement has begun.
Where contractors accept fixed prices or aggressive completion dates, those changes can erode margin quickly. The commercial attraction of a £1bn development pipeline therefore depends on who carries inflation, equipment delay, design development, labour productivity, and commissioning risk rather than headline project value alone.
US scale offers Balfour Beatty an established route
Balfour Beatty already has a sizeable US buildings business with regional teams, subcontractor relationships, procurement processes, and repeat customers. Expanding data centre work inside that platform is different from building a new mission critical operation across several European jurisdictions.
Europe is fragmented by national planning systems, electrical standards, utility structures, labour markets, procurement models, and construction practices. A delivery method established in Virginia or the US Pacific Northwest cannot simply be transplanted into London, Frankfurt, Madrid, or Helsinki.
Even within Europe, individual data centre markets behave differently. London developers face constrained power and expensive land, Frankfurt projects are strongly influenced by grid access and sustainability requirements, while southern European sites introduce hotter design conditions and, in some locations, greater water pressure.
Contractors entering those markets need enough repeat work to justify specialist teams without becoming dependent on one customer or accepting poor terms simply to establish a position.
Balfour Beatty’s US Buildings business already operates at relatively thin construction margins, making further margin compression unattractive. High project value can obscure how little room remains after subcontractor costs, design changes, procurement problems, and programme delay are absorbed.
Contractor capacity belongs in the buildability equation
European data centre development is usually framed around land, planning, and power, but the ability to secure experienced contractors and specialist MEP teams can also determine whether capacity arrives on schedule.
AI facilities add another layer of complexity. Higher rack densities require larger electrical feeds and increasingly sophisticated liquid cooling, while customers often want capacity quickly because accelerator hardware and AI service demand move on shorter commercial cycles than conventional building programmes.
Compressing construction can increase overtime, procurement pressure, rework risk, and the number of activities taking place concurrently. Unless developers retain some of that risk or pay enough to compensate the supply chain, experienced contractors can decide that other infrastructure sectors offer better returns.
Balfour Beatty can also capture parts of the data centre expansion without constructing the facilities themselves. The group is targeting UK energy infrastructure, and the transmission lines, substations, cables, and grid reinforcement needed for new large loads form a substantial construction market in their own right.
That division of work may become more common as the sector expands. Some contractors will specialise in mission critical buildings, others in grid and civil infrastructure, while electrical and mechanical specialists take larger packages within the campus.
Balfour Beatty’s US growth shows that it is prepared to build data centres where the commercial structure works. Its refusal to pursue the same strategy in Europe places pricing discipline against a market that is otherwise characterised by large pipelines and strong demand forecasts.
If other major contractors reach similar conclusions, developers will face a straightforward choice: accept a narrower delivery pool, change procurement structures, retain more programme risk, or pay more for the construction capacity needed to turn announced megawatts into operating facilities.
