The industrial E&I subset — plant wiring, instrumentation loops, control panels, motor-control centers — hides inside a category dominated by commercial construction electricians. Industry codes cannot tell them apart.
Case-study pages can, and we read every one of them, at census scale, against your thesis.
Electrical contracting is enormous. The industrial slice a services thesis wants is small, specific, and mislabeled almost everywhere.
NAICS gives every electrical contractor one code — the firm pulling wire on apartment build-outs and the one calibrating instrumentation loops in a chemical plant look identical in every deal-team tool. For an industrial-services thesis they could hardly be more different businesses.
State-issued, often county-endorsed. The qualifying license typically belongs to a founder — anchoring firms geographically and keeping them independent.
Plants do not hand instrumentation work to strangers. A miswired loop or botched MCC bucket is downtime measured in production hours.
The deep-extraction pass classifies each candidate into the structure below — from documented evidence, never from the label the company inherited.
Ongoing maintenance electricians for manufacturing sites: MCC work, VFD installation, power distribution, lighting retrofits inside operating facilities. Recurrence comes from site relationships and on-call arrangements rather than formal contracts — visible in service-page language.
Loop checks, transmitter calibration, analyzer maintenance, DCS and PLC support. The scarcest and most defensible capability set in the category — and the least keyword-visible, because these firms describe processes, not trades.
Control-panel fabrication, often UL 508A certified, frequently paired with field services. The UL 508A listing is a one-line website claim that instantly separates a real panel shop from a reseller — we quote it exactly when present.
Commercial build-out and new-construction firms. Often excellent companies — and usually documented exclusions under an industrial-services thesis, because bid-cycle project revenue is a different underwriting problem than plant-service recurrence.
We do not search for E&I contractors. We start from everything and exclude our way to them — the direction of travel changes what you find.
Profile databases assemble lists first and answer questions second — their answer to "who does industrial E&I in these twelve states?" is bounded by who they previously indexed. We invert that: start from the full classified web and exclude our way to the answer.
Extracted alongside the rest of the 15-signal framework, but these carry the weight in this vertical.
The decisive signal. Industrial E&I firms prove themselves through named manufacturers in case studies, water-utility projects, food-plant washdown-rated installations, refinery shutdown support. We extract end markets only from project pages, customer lists, and substantive industry pages — never from a homepage adjective. A contractor claiming "industrial" with nothing but office build-outs in its gallery is classified by the gallery.
Licenses and certifications are published because they win work: state master electrical licenses, UL 508A panel-shop listings, hazardous-area competencies, safety-scheme memberships, stated EMR figures, NFPA 70E programs. Each captured as exact claim text from the source page. Individually small, collectively decisive — and absence is informative too.
The split runs service-versus-construction and changes acquisition economics utterly. Service-led firms show on-call language and troubleshooting depth; construction-led firms show project galleries and GC relationships. Most are a blend, so we record the documented weight of each — a 70/30 service-led firm and a 30/70 construction-led firm are different theses.
A firm with named service managers, project managers, and estimators is a different acquisition than a one-truck operation with a founder's phone number. We count who is actually named and in what functions — the best website-level proxy for whether the business runs on systems or on one person's pager. Combined with founder-association language, it feeds the Transition Context score without inferring anything about private individuals.
Named relationships position a firm in the vertical's ecosystem: authorized integrator status with automation vendors, distributor partnerships, trade-association memberships. Quoted from partner pages and badge walls, they matter twice — as capability evidence and as a map of which vendor ecosystems a platform would be buying into. An acquirer consolidating around one control-system ecosystem can filter the delivered universe on exactly this field.
A composite narrative, faithful to how real runs unfold — with the arithmetic that surprises buyers the first time.
Take a single industrial metro — the triage pass surfaces 400 live domains carrying electrical-services vocabulary. Here is how the deep pass resolves them.
Calibration language, loop-check services, panel-shop listings
Second-generation language or founder named as license-qualifying party
"Automation services" or "plant services" firms with full E&I practices revealed in deep pages
Skeptical readers raise the same three points. They are correct, and the method is built around them.
They do — in adjectives. They understate in structure: a contractor can call itself industrial, but it cannot fake a UL 508A listing, a plant-floor project gallery, or ten years of shutdown case studies. We lean on structural evidence and quote it verbatim — inflation shows up as absence of substance, which is itself recorded. Where a site offers too little, the company is flagged insufficient-evidence rather than guessed.
Correct, and we refuse to pretend otherwise. No revenue estimates, no EBITDA bands, no valuation guesses — a private E&I contractor's website carries none of that reliably. The screen tells you who exists, what they demonstrably do, and how they fit your thesis. Sizing them is what your first conversation and your quality of earnings are for.
Also correct. We never claim to detect readiness to transact — no website signal supports it. What evidence does change is outreach quality: referencing a firm's actual instrumentation practice and named end markets converts at a different rate than a mail-merged industry blast. The list is the beginning of your process, done properly.
Structured records import directly into whatever CRM your team runs. E&I mandates typically sort the universe three ways on arrival.
Immediate partner review — calibration depth, controls expertise, panel-shop capability documented
Standard outreach cadence — on-call language, plant-maintenance relationships, MCC and VFD work
Reference tier — keeps future channel checks efficient; each carries its quoted exclusion reason
| If your constraint is… | Then… |
|---|---|
| Speed to a first credible shortlist | A proof project (from €4,900) screens a defined slice end to end in days — judge the classification quality on companies you can verify yourself |
| Confidence the universe is complete | Census-based screening is the only approach that can state coverage: every active domain in the category was evaluated, and exclusions are documented, not silent |
| Avoiding wasted outreach | Evidence-quoted classifications remove the construction-electrician false positives before they reach your sequences — the largest single noise source in this category |
| A thesis still being refined | Included ICP re-runs mean the universe is built once and re-questioned freely — refinement stops costing a restart |
| Board-defensible process | Verbatim quotes and source URLs on every inclusion and exclusion give an IC memo its audit trail |
One email with the thesis as you have it. The specimen arrives the same day; the scoped proposal follows within the week.
Request the specimen report