Per-MW pricing, regional variance, and cost drivers for owners scoping hyperscale & AI builds.
Salary benchmarks across the 14 mission-critical disciplines.
Ask what the candidate personally owned, what changed, and who can verify the result. A data center project title alone doesn’t prove readiness to lead the work.
Focus interview questions on six areas:
Rate each area as Proven, Partial, Unverified, or Red flag. Check claims through redacted records or approved references. Missing paperwork calls for follow-up - not an automatic rejection - but unsafe judgment or false ownership claims should not be offset by a high overall score.
Data Center PM Interview: Six Evidence Checks
Start with the candidate’s largest project. Verify three decisions they personally owned, and connect each to delivered MW and the delivery method. Confirm whether the work was a greenfield build, expansion, fit-out, or retrofit. Ask which option they rejected, who approved the decision, and which milestone changed.
Test reliability knowledge through sequencing - not definitions alone. Uptime Institute distinguishes Tier III’s concurrent maintainability from Tier IV’s fault tolerance. Equipment count alone does not prove a Tier level.[2][3] Ask how the candidate sequenced electrical and cooling paths, which paths needed independent acceptance, and how temporary power and cooling affected redundancy. What conditions had to be met before energization? Verify who owned those decisions and whether the milestone held.
Use these same project facts to check how the candidate mitigated schedule risks when equipment deliveries or interface work slipped.
How have you recovered a delayed critical-path equipment delivery?
Require original and revised delivery dates, affected schedule activities, available float, trade-offs, and actual days recovered. A strong answer follows the full path from design release to turnover. Adding labor is not recovery unless executable work is available. Check access, materials, inspections, and testing constraints before accepting an overtime plan.
Ask what the candidate would do in the first week after a milestone slips. On Day 1, confirm supplier status and notice rights. On Days 2–3, update field progress, remaining durations, schedule logic, and milestone impact. On Days 4–5, issue approved options, vendor commitments, assigned owners, and a dated recovery forecast - not an unexplained replacement of the baseline.
Use the schedule review to check whether the candidate can protect milestones without sacrificing commissioning readiness. Ask how consuming float changes the forecast. Strong candidates distinguish total float from free float, identify when another path becomes critical, and explain who receives an escalation and what decision is needed.
Verify forecasted versus actual milestone dates. Did recovery preserve inspections and testing, or just push unfinished work into turnover? Carry that forecast review into the next check: which MEP interfaces, vendors, and test gates were at risk?
Once schedule control is clear, check whether the candidate can coordinate MEP work to protect energization and testing. In a data center, a single unresolved interface can delay energization or stop testing.
Ask candidates to walk through a tight data-hall interface they resolved before installation. Which systems conflicted? Did field measurements match the BIM model? Who approved the fix? A clash-free model alone isn't enough. Layouts must also account for installation sequence, maintenance access, and testing access.
Require a clear approval trail. The GC coordinates the fix and enforces contract requirements, while the engineer of record and owner retain technical approval responsibility. Check the RFI, submittal, drawing, and change-control records that support the solution.
Ask when commissioning staff and operators first reviewed service access, test points, controls submittals, and sequences of operation. Commissioning starts during programming and design - not at final inspection.[1] A candidate who defers controls until commissioning or accepts unapproved field changes needs closer scrutiny.
Once the layout checks out, verify that trades and vendors can build and test it without rework.
Ask how candidates prequalify mission-critical trades beyond price and past relationships. Require comparable-project references, staffing depth, financial capacity, fabrication resources, quality records, and commissioning support. Familiarity matters less than the ability to support the installation sequence, startup, and recovery.
Then ask about owner-furnished equipment. Who handles receiving inspections, storage, rigging, installation, controls connections, vendor startup support, and warranty requirements? Each responsibility needs a clear owner.
Ask what the candidate did in the first 48 hours after a trade fell behind. Look for a verified cause, a capacity check, named owners, dated targets, required notices, escalation, and proof of closure. Score documented decisions - not blame.
Use those same records to assess whether the candidate can also hold cost and turnover targets later.
At 70% complete, ask candidates how they forecast final cost. Look for an estimate-at-completion method that reconciles the budget, approved changes, commitments, actuals, remaining work, escalation, contingency, and risk. For data center projects, it should also account for unfinished testing, corrective work, vendor return trips, and retesting. A data center PM should show how recovery decisions protected both the budget and turnover.
Require approved, pending, disputed, and potential changes to be tracked separately, with cost, schedule impact, owner, and decision date for each. Approved changes should update the budget. Pending changes should stay visible without being counted as approved budget. Request redacted monthly cost reports, estimate-at-completion revisions, and change logs. Ask candidates to explain one variance, then check references to verify their personal ownership. The same decisions should also appear in the turnover forecast.
Ask how they weighed overtime, expedited freight, added supervision, temporary equipment, resequencing, or alternate suppliers against the cost and schedule effects of delay. Require the assumptions, approval record, and actual milestone outcome. Watch for red flags: invoice-based forecasts, unsupported progress estimates, and answers such as:
“recover first, price later”
Use the cost forecast to check energization and turnover readiness. Ask candidates to walk through how they coordinated energization and integrated systems testing (IST). Their answer should distinguish equipment startup, pre-functional checks, functional testing, and IST under normal and failure conditions.
Require clear go/no-go criteria covering approved procedures, qualified personnel, completed prerequisites, calibrated instruments, switching plans, lockout/tagout boundaries, and applicable arc-flash controls. NFPA 70E guidance emphasizes identifying all sources, opening disconnects, applying lockout/tagout, and verifying absence of voltage with a properly rated instrument when establishing an electrically safe work condition.[5][6]
For a failed test, require the deficiency record, assigned corrective owner, root cause, cost exposure, and successful retest - not just a closed work order. Then inspect the readiness tracker for required test results, deficiency closure, operator training, as-builts, O&M manuals, warranties, and owner signoff.
ASHRAE’s guidance ties handover to completed IST, resolved defects, delivered documentation, and owner acceptance of the turnover procedure.[4] Treat commissioning as proof that systems meet the owner’s operating needs, not as paperwork. Unresolved safety issues or issues affecting system operation are hiring red flags.
These questions are designed to help GCs navigate the complexities of data center construction and ensure their project managers are prepared for mission-critical demands.
After reviewing schedule, MEP, cost, and commissioning performance, close the interview by scoring every candidate against the same evidence.
Put verified outcomes, decision quality, and demonstrated ownership ahead of project labels. Focus on the decisions candidates personally owned and the measurable results they delivered.
Rate every candidate using the same four options: Proven, Partial, Unverified, and Red flag. Record the supporting evidence immediately after each interview, then compare candidates side by side using evidence from their answers.[7][8]
Set minimum requirements before interviews. Address gaps with structured follow-up questions, work samples, and authorized references. Don't let an overall score hide unsafe judgment or unsupported ownership claims. Ask every reference the same role-related questions and document the evidence consistently.[7][9]
Before making a decision, document each rating, note what still needs verification, and flag any safety or integrity concerns.
Prioritize candidates with high-reliability experience over those with general commercial construction backgrounds. The most successful transitions come from semiconductor fabrication, pharmaceutical manufacturing, healthcare, or large-scale industrial construction.
Look for documented experience managing complex mechanical, electrical, and plumbing (MEP) coordination, phased delivery, and rigorous commissioning. Candidates must also have direct experience with integrated master scheduling and mission-critical systems where downtime isn't an option.
Ask for redacted or anonymized materials that show the candidate’s scope of work and authority to make decisions. Nonconfidential summaries are fine. They should cover the facility type, IT capacity in megawatts, project value in U.S. dollars, and delivery model. Leave out restricted configurations, proprietary data, security-sensitive details, customer names, and contract terms.
Assess technical expertise with scenario-based questions about test sequencing or resolving deficiencies. Then confirm the candidate’s performance through firsthand references from project executives or partners.
Request a project record for each relevant assignment. Each record should include the facility type, ground-up or retrofit scope, project value (USD), IT load (MW), gross square footage, delivery model, and team size. Ask candidates to distinguish the phases they personally led from those they supported so you can verify their experience and decision-making authority.
Accept redacted schedules, commissioning logs, and closeout records when confidentiality is required. These samples should show how the candidate tracked procurement, energization, functional testing, and turnover milestones, along with how they managed MEP interfaces within project constraints.
