Data Center Work Is Booming. Here’s What It Actually Takes to Get In.

The Data Center Construction Boom Is Real and Right Now

The scale of what is being built is difficult to overstate. Approximately 12 gigawatts of data center capacity is expected to come online in the US this year alone, nearly triple the capacity built in the prior year, according to Sightline Climate.^[3] Gartner estimates more than $475 billion in total global data center spending for the current year, and McKinsey projects that AI-driven infrastructure investment could require as much as $5.2 trillion globally by 2030.^[4]

The geographic spread of this demand is also widening well beyond established markets. According to JLL, 64 percent of new data center capacity under construction in North America is now located in frontier markets outside traditional hubs like Northern Virginia and Silicon Valley.^[2] In Canada, Alberta has emerged as one of the continent’s most active emerging data center markets, with Microsoft, AWS, and other hyperscalers announcing or expanding facilities in the province, drawn by available land, cooler climate reducing cooling costs, and power infrastructure that is comparatively less constrained than major US grid markets. As DPR Construction’s most recent Market Conditions Report notes, this frontier expansion means finding qualified mechanical, electrical, and structural trades is now nearly as difficult as securing the megawatts required to power the facilities.^[5]

In fact, some hiring roles on these projects that previously took eight weeks to fill are now taking four months or more.^[5] McKinsey estimates the US will need 130,000 additional electricians by 2030 to meet AI infrastructure demand alone.^[6] Accordingly, the contractors who are operationally positioned to bid and execute this work are in a position most rarely occupy: more work available than the market can absorb.

What Data Center GCs Actually Evaluate at Prequalification

Most mechanical contractors approach data center prequalification the same way they approach any commercial project: project history, bonding capacity, safety record, insurance certificates. Those things are necessary. They are not, however, sufficient.

Across hyperscaler GCs, the prequalification evaluation reduces to five operational and financial items every subcontractor is scored. The first three are widely known. The next two are where most mechanical and specialty contractors quietly lose the evaluation, one operational, one financial. Here is what each one actually means in practice.

1. WIP schedule of values
   Available on demand, not pulled only at month-end close. The Construction Financial Management Association treats the WIP schedule as a live financial management tool rather than a reporting artifact, which is the bar hyperscaler GCs now expect. A contractor who can only produce a WIP schedule at billing milestones is signalling to a data center GC that their financial visibility lags their project reality.
2. Real-time cost-to-complete
   Estimate-to-complete rolled up across every active project, not just the one being bid. On a portfolio of concurrent jobs, this is the number that tells the GC whether you can take on more work without absorbing margin elsewhere. Most mid-market MEP contractors can produce a cost-to-complete for a single project. Producing it across all active scopes simultaneously is the capability gap that separates contractors who get invited back from those who do not.
3. AIA G702/G703 billing compliance
   The standard Application and Certificate for Payment format every hyperscaler GC expects on progress billing. Submitting a non-compliant pay app format on a data center project is one of the fastest ways to be quietly removed from the next round of bids. This is not a documentation preference. It is a baseline operational requirement, and contractors running billing out of QuickBooks or disconnected systems routinely fail it without realising the damage until they stop receiving invitations.
4. Change order log with approval status
   Every scope change logged, priced, and tracked against approval status before additional work proceeds. This is the single most common reason specialty contractors are cut from data center bid lists. The dollar value of unapproved or untracked change orders accumulates faster on mission critical work than on standard commercial projects. GCs review the change order log specifically to assess whether a subcontractor will manage scope cleanly or generate disputes downstream. A contractor whose change order process is informal, meaning scope changes are tracked in emails or spreadsheets rather than a documented approval workflow, is a liability on a project where a single disputed change order can run to seven figures.
5. Financial stability ratios and bonding capacity
   Working capital, debt-to-equity, current ratio, and aggregate bonding capacity reviewed against the same Character, Capacity, and Capital framework hyperscaler GCs adopt from their bonding companies.^[7] Most specialty contractors do not realize their financial ratios are being scored in parallel with their operational fit during the same evaluation. A contractor with strong crews and solid references who is underweight on working capital relative to the project size will be quietly passed over. The financial stability review is not a formality. It is the filter that runs concurrently with everything else and that most contractors never see coming.

The implication of this five-item framework is important. Three of the five criteria, WIP reporting, billing compliance, and real-time cost visibility, are software problems as much as they are process problems. A contractor whose system cannot produce these outputs in the required format cannot compensate with experience or relationships. The technical evaluation never begins.

Why Standard Tools Break Down on Data Center Projects

Many mechanical contractors pursuing their first data center projects encounter the same problem at the same point: the tools that have served them well on standard commercial work are not built for this level of complexity.

The gap is typically invisible until a prequalification surfaces and exposes it. A contractor submits a package that looks complete. The GC requests a current WIP schedule of values on demand. The contractor’s system cannot produce one outside of month-end close. The evaluation ends before the technical review begins.

Beyond prequalification, the most common operational gaps that hurt mechanical contractors on data center projects include:

• No real-time cost visibility: On a project where scope changes weekly, margin erosion happens fast. Without job costs updating in real time from the field, decisions are made on data that is already outdated.
• Disconnected field and office systems: When field teams use one platform and accounting uses another, the result is duplicate data entry, delayed reporting, and billing that does not reflect actual project conditions.
• Reactive procurement management: Given transformer and switchgear lead times measured in years, reactive purchasing is no longer viable on data center work. Procurement must be integrated with the project schedule and flagged against delivery risk from the outset.
• No unified construction and service view: Many mechanical contractors carry both a construction division and a service division. Without a platform that manages both from the same system, the business loses visibility when a data center client transitions from construction to ongoing maintenance.

What Winning Data Center Contractors Do Differently

Despite the challenging environment, a segment of mechanical contractors is successfully delivering data center work and building repeat relationships with hyperscaler general contractors. Generally speaking, they share operational characteristics that set them apart.

They treat procurement as a strategic function

Rather than purchasing equipment reactively, high-performing data center contractors build procurement planning into the project timeline from day one. Specifically, they track long-lead items against delivery risk, maintain relationships with alternative suppliers across multiple geographies, and escalate procurement issues before they become schedule impacts. Given Wood Mackenzie’s documented long lead times for key equipment like power transformers and generator step-up units,^[8] procurement discipline is now a competitive differentiator on data center work, not a back-office function.

They run integrated financials and field operations

Rather than maintaining separate tools for project management and accounting, successful contractors use a unified platform that connects field labour, job costing, subcontract management, procurement, and billing in real time. In effect, this eliminates the lag between what happens on site and what leadership sees on the P&L, giving the business the visibility to act before margin is lost rather than after.

They manage change orders as a core discipline

On data center projects, scope changes are frequent and the dollar values are significant. Winning contractors treat every change as a formal process: logged, priced, and tracked against the approved contract before additional work proceeds. Subsequently, this discipline protects margin and reduces billing disputes with GCs, which directly affects whether a contractor gets invited to the next project.

They position for long-term service relationships

Data center operators need ongoing maintenance partners after commissioning. Mechanical contractors who can demonstrate a unified construction-and-service capability, managed from the same platform, are more attractive to hyperscaler clients looking for long-term reliability. Moreover, those service contracts represent recurring revenue that stabilizes the business between construction cycles.