Where the work is, the Phoenix area
U.S. colocation inventory reached 29.0 GW at the end of Q1 2026, up 22% quarter over quarter and 48% year over year. The binding constraint on that growth is a specialist labor bottleneck, not generic construction labor: qualified electricians, mechanical and controls specialists, commissioning teams, and high-voltage field talent.
Arizona holds roughly 2.4 GW of deployed data center capacity, 1.3 GW under construction and 4.1 GW planned, with a 12.8% CAGR projected to 2030. The work concentrates in two parts of the Phoenix metro: the West Valley (Goodyear, El Mirage, Buckeye, Glendale) and the East Valley (Mesa, Chandler, Queen Creek), with Tucson emerging. Committed capital in the Phoenix area, per public announcements: Google's $1B, three-phase, 187-acre project in Mesa; Microsoft's 576.5 acres across three sites in Goodyear and El Mirage; Compass's 225 acres and 350 MW in Goodyear; Meta's $1B, 2.5 million square foot project in Mesa, built by DPR; EdgeCore's $1.9B campus in Mesa, built by Holder; Novva's 165-acre campus in Mesa with a plan to reach 1 GW by 2027; and Edged Energy's $70M, 210,000 square foot, 36 MW waterless-cooling building in Mesa, built by Haydon, a Phoenix general contractor.
Utility reality: SRP's data center load hit 441 MW, 5.1% of system peak, up 425% in six years. APS, SRP and TEP interconnect queues are real but not saturated. Power coordination is a discipline to build, not an afterthought.
The mini and edge segment is the open lane underneath the hyperscale campuses: sites from 50 kW to 5 MW, plus mid-size single buildings into the tens of MW, modular and prefab-friendly, placed closer to users than a hyperscale campus can be. Novva's Mesa campus already serves tenants from 250 kW to 30 MW, and Edged Energy's 36 MW single building was won by Haydon, a Phoenix general contractor, not a national one. Prefabricated modular delivery, factory-built pods that arrive prewired and pretested, is the construction model for this lane, cutting on-site cabling and testing by up to about 70%; the long pole at every site is utility power, not the building.
Arizona's Computer Data Center Program (A.R.S. section 41-1519) exempts qualifying equipment from state, county and local transaction-privilege and use tax for 10 to 20 years, on a threshold of $50M invested in Maricopa or Pima County ($25M elsewhere) within five years. For a mini-site portfolio, the move is to aggregate multiple sites under one Arizona Commerce Authority certification.
Licensing and bonding
The plan works through the state contractor licensing stack, from a general commercial license up through electrical, mechanical, low-voltage and other trade classifications, each carried by a qualifying party who passes the required trade and business exams. The electrical classification is the margin and the schedule control. LÏEF/Armstrong already holds an operating Arizona contractor license, bonding history and bid infrastructure, a running start that gets verified and confirmed with the state registrar in the plan's first week.
The electrical qualifying party, a master electrician with mission-critical experience, is the single most important hire in the plan. The plan runs two paths at once: recruiting a senior hire from the regional mission-critical electrical bench or the local electrical union hall, or acquiring a licensed electrical subcontractor and retaining its qualifying party under a fixed-term agreement. Arizona's licensing reciprocity with neighboring states is a trade-exam waiver only, never a license transfer, so the plan licenses Arizona first and qualifies any second state independently when expansion is real.
Bonding capacity, not the license itself, is the real ceiling on job size, and it grows with balance sheet and track record. Early jobs structure around that ceiling by taking electrical-subcontractor scope or joint-venturing under an incumbent general contractor's bond while the proposed division's own bonding grows, building from the existing LÏEF/Armstrong surety relationship. A company-level safety program, owner prequalification packages, uptime-tier delivery fluency, individual arc-flash and commissioning certifications, and manufacturer certifications on switchgear, generators and cooling equipment round out what would get the proposed division onto owner shortlists.
The team
A seven-person founding core scales in phases to an electrical bench and eventually a merchant platform. Existing LÏEF labor relationships transfer to the civil, structural and shell side of the work; the mission-critical layers, electrical, cooling, low-voltage and controls, commissioning, have to be recruited, poached, grown through apprenticeship, or bought.
Three hires carry the plan. The electrical qualifying party, the license rides on this person, and the recruiting pitch is equity and ground-floor ownership the larger employers structurally cannot offer. A mission-critical project executive who has delivered data centers end to end, so LÏEF's principal would own relationships and capital while the project executive would own delivery. And a commissioning lead, fractional at first, full-time once the line proves, who carries the credentials that open the commissioning revenue line and make every bid smarter.
Cash hires for the qualifying party and the project executive fire only after an anchor commitment is papered; fractional and consulting roles, commissioning, interconnect coordination, safety, start immediately at low cost. Later phases add an electrical crew structure, a mechanical sub-partnership before an in-house crew is justified, an apprenticeship pipeline as the durable answer to a tightening wage market, and eventually an in-house commissioning and energization team on its own books, a modular supply-chain function, and licensing in additional states if the pipeline demands it. Equity participation for the first hires is the plan's answer to competing with much larger employers for the same scarce talent, alongside a standing training and certification budget.
Buy the team
Acquiring a small, licensed Arizona electrical subcontractor is the plan's strongest move: one transaction can deliver the license, the qualifying party, the crews and a revenue base at once, collapsing a hiring timeline that would otherwise run the better part of a year in the tightest electrical labor market in memory, since the trades are full of succession-motivated owners with no exit plan.
The target profile is a commercial electrical contractor with a clean licensing history, a Phoenix-metro base, and an owner 55 or older looking for an exit; mechanical and low-voltage subcontractors are secondary targets for the same logic, with the low-voltage and building-management-systems trade an inexpensive but outsized strategic asset for the controls and commissioning lines.
Deal terms are built around retention. The qualifying party signs a fixed-term employment and qualifying-party agreement, since the license rides on this person and there is no deal without it. Key foremen get retention bonuses, because the crews are the asset and the deal is worthless if they walk. Work-in-progress schedules and bonding history get diligenced hard, since half-finished, underwater jobs are the classic trap in trade-subcontractor acquisitions, and customer contract assignability gets confirmed before close. The process runs on a dated track: building a long list from license rolls and industry association membership, quiet succession-framed outreach to a short list, then a letter of intent, diligence and a close with counsel. The acquired business keeps running under its own name and brand after close, cash flow uninterrupted, while the data center program layers on top and the first cross-staffed project pairs the acquired crews with the mission-critical project executive.
Services and revenue
Four lines, one team: the pitch to owners would be one team from dirt to energization.
General contracting and design-build delivery for mini and mid-size projects, modular-first, pod placement, utility tie-ins and site integration for prefabricated deployments, priced as a fee on construction value, targeting jobs below the megaproject general contractors and above the generic commercial contractors who lack mission-critical credibility.
Electrical self-perform: medium-voltage distribution, switchgear, uninterruptible power and battery systems, generator plants, grounding, feeders and terminations, both inside the proposed division's own general contractor jobs for margin capture and as a specialty subcontractor on other builders' projects for market entry.
Commissioning and energization for hire: independent commissioning work on projects LÏEF did not build, since owners must hire independent commissioning and every data center built by anyone in Arizona is a prospect. The proposed division would not commission its own general contractor work; that separation would be the product, and the work is people and instruments with no materials risk, counter-cyclical to the construction cycle.
Interconnect advisory, acting as an owner's representative for power: site feasibility, utility interconnect applications and queue management, and long-lead procurement strategy for transformers and switchgear. This line starts first, in month one, on an anchor's site pipeline, builds the utility relationships every later job needs, and qualifies an anchor's own sites honestly, including which sites will never get power.
The sequencing runs interconnect advisory first, in month one; electrical self-perform entering as a specialty subcontractor within the first several months; the first full general-contractor job landing between months nine and eighteen, likely alongside an established general contractor for prequalification cover while the proposed division's own bonding grows; and by the second year all four lines running, with commissioning eventually standing up on its own dedicated team.
What could kill this
Ten concerns were tested, two of them critical, both carrying explicit gates rather than hope. Anchor capital is unverified: the pipeline thesis rests on a raise that has not yet been papered, resolved through standard reciprocal diligence, and founding cash hires wait on it; without verifiable capital by a set date, the plan pivots to a merchant-only thesis or is parked. Power gates every site: transformer lead times run long nationally, only a fraction of announced capacity is actually under construction, and a site without secured power is a stalled site with idle crews, so every site gets a power-feasibility verdict before any construction commitment; zero power applications in the pipeline means restructuring to an advisory-first posture.
The rest of the matrix is managed, not ignored: qualifying-party dependency, since the electrical license rides on one person, answered by targeting two qualifying parties within the first year and employment agreements with notice provisions; unconfirmed licensing classification assumptions, cheap to resolve in a first-week consult and embarrassing to skip; a bonding ceiling that starts small and needs balance-sheet support to grow, answered by early jobs structured as subcontractor scope or joint ventures and by a partner balance sheet under the platform structure; a labor-cost premium that risks a poaching war, answered with equity participation and an apprenticeship pipeline as the long-term hedge; water and growth-management politics in Arizona, answered by waterless and air-cooled design competence and by mini-footprint sites drawing less heat than hyperscale; concentration risk from launching around one client's pipeline, answered by building merchant revenue lines from day one; management bandwidth, since the principal runs multiple ventures, answered by hiring a project executive who owns delivery so the principal owns relationships and capital only, with the proposed division not launching without that hire; and the risk of larger general contractors moving downmarket in a slowdown, which speed, local cost structure and self-perform electrical economics are built to withstand.
If this plan failed by mid-2027, the most likely causes, in order: anchor capital never verified and the proposed division burned overhead waiting; a site stalled on interconnect for an extended period with crews idle; or the electrical qualifying party leaving and the license gapping mid-project. All three have explicit gates built into the plan, and a full premortem runs before any engagement signs.
The ninety-day plan
A dated sequence from a first licensing consult to a first project pursuit, with hard gates so overhead never outruns commitment.
The first two weeks open the licensing and bonding conversations with a qualifying-agent consult, an insurance broker conversation, a quiet search for electrical qualifying-party candidates, a first acquisition long list, and preparation for a deep-dive meeting with the anchor, including wargaming the conversation and a premortem of the proposed division itself. That meeting walks the plan together, exchanges reciprocal diligence, presents the structure options and states a lean, offers Arizona's data-center tax program as a show of expertise, and leaves with a dated follow-up and a decision gate of thirty days at the outside.
If the anchor signals real, the following weeks paper the anchor commitment at a letter-of-intent level, decide the electrical licensing path (hire or buy), begin qualifying-party and acquisition outreach in earnest, start interconnect advisory work on the anchor's site list, open a search for the project executive, and begin a second version of the financial model with Common Ground.
The final stretch of the ninety days aims to secure the electrical classification, hire the project executive, retain a fractional commissioning lead, stand up the safety program, and pursue two projects in parallel, an anchor site if it qualifies on power and one third-party pursuit, so the proposed division would never be single-client from day one, closing with a go or no-go review against the two gates.
Standing disciplines throughout: no founding cash hires before the anchor papers, every site gets a power-feasibility verdict before any construction commitment, and every item on the plan expires in two weeks, shipped or killed.
The engagement
Three ways to structure the anchor relationship, a menu rather than a fixed proposal, with the right structure depending on what reciprocal diligence shows on both sides.
A captive-builder structure would have the proposed division build only the anchor's sites, predictable but capped, and carries concentration risk if the pipeline goes quiet between projects; it fits if the anchor's capital verifies for several funded sites and a standby arrangement covers the gaps between them. An anchor-plus-merchant structure would have the anchor commit a pipeline at preferred terms and priority scheduling while the proposed division would also serve the broader Arizona market, balancing anchor base load against full merchant upside; it fits once capital verifies for a first site alongside papered multi-site intent. A platform joint venture would have the anchor invest directly in the proposed division itself, funding the acquisition of an electrical subcontractor and the proposed division's stand-up, in exchange for equity and priority, plus the option to co-invest per site; it fits once capital is fully verified and both sides want alignment beyond client status, properly papered.
Our lean is the anchor-plus-merchant structure, with the platform joint venture as a natural evolution once one project of working history is behind both sides, on a pre-agreed option framework: a builder that builds for only one client dies between that client's projects, and an anchor agreement delivers priority and pricing without paying for idle time.
Every structure would run on open-book estimating so preferred terms are auditable rather than asserted; on ordinary construction contracts rather than securities, with any capital raising routed to a licensed broker-dealer platform and never through the proposed division itself; on honest power verdicts, even when that news kills near-term construction revenue; and on reciprocal diligence, with fund documents and site control opened on one side and the proposed division's plan, model and licensing path opened on the other, both before either side commits. Arizona's data-center tax incentive program, which rewards aggregating multiple sites under one certification, comes free as part of the plan and is offered as a show of expertise regardless of which structure gets chosen.
The June plan's next step was a deep-dive meeting to walk this plan together and exchange diligence lists, followed by a thirty-day window to paper an anchor commitment at the letter-of-intent level, or for both sides to walk away clean. From there the ninety-day plan executes: licensing, the electrical path, interconnect advisory on the pipeline, and first-project preconstruction.
What it produced
A provisional finding, as of June 2026, with two gates still open: an anchor capital commitment that has not yet been verified, and a licensing classification that has not yet been confirmed with the state registrar. This is an exploration, not a decision to launch. The June plan's next step was a deep-dive meeting to walk the analysis together, exchange reciprocal diligence, and set a thirty-day decision gate. Founding cash hires do not fire until the anchor commitment is papered, and no site gets a construction commitment before a power-feasibility verdict.
A slice of the project list
A few related projects.
- Canyon Corporate: commercial construction and development (2026)
- Village San Juan: residential construction (2026)
- Soft-Story Retrofits: seismic retrofit program (2026)