Data Centre Buildout Hits a Talent Wall: Kelly Warns of Acute Skilled-Worker Shortage
The News
Data centre development has created acute labour shortages in construction, utilities and telecommunications, driving up wages and delaying projects, according to a report from Kelly Services cross-referenced with US Bureau of Labour Statistics data. Data centre employment is expected to reach 650,000 positions by 2026, a 30% increase from 2023, with related construction jobs exceeding 180,000 through 2028. Yet 25% of data centre personnel are hired away by competing hyperscalers and operators. National midpoint salaries start at $175,000 for AI engineers and $178,000 for data centre operators, varying by 46 percentage points across the US. Kelly attributes the primary controllable cause of delays to workforce planning failure rather than power, land, or equipment constraints.
Why It Matters
The data centre story is the sharpest expression of a problem TA leaders across sectors are seeing in softer form: AI infrastructure investment is outpacing the talent supply needed to build and run it, and the competition for skilled workers is now pulling talent out of adjacent industries. For TA leaders in technology, engineering, and infrastructure, the signals are concrete — salaries at $175K+ midpoints, 25% annual poaching rates, and time-to-hire for skilled trades now longer than for knowledge workers. The Kelly recommendation to recruit from adjacent sectors (telecoms, utility grid, industrial HVAC) is a skills-based-hiring case study in miniature, and worth examining even if your organisation has no data centre roles to fill.
Taking Action
All action is situated in your corporate context. These are starting points, not prescriptions — choose what fits.
Connect
Who to talk with in your enterprise
Your workforce planning lead: ask whether your organisation's AI or infrastructure hiring plans have been stress-tested against a 25% attrition rate for the roles most exposed to external poaching. If not, the headcount plan is fiction. Your head of talent acquisition for technical roles: discuss whether you have identified adjacent-sector talent pools you could recruit from if your primary sources dry up.
Inspect
What to take a closer look at
Pull the 12-month attrition rate for your most specialised technical roles — the ones where external demand is highest. If it exceeds 15%, compare your retention package (pay, mobility, development) against what data centre operators are now offering. If you cannot identify at least two adjacent sectors you could source from for each critical role, your hiring pipeline has a single point of failure.
Reflect
What to think about more deeply
When the binding constraint on your AI strategy is not compute power but the people who build and run the compute, is your talent acquisition function resourced as if that were true?