The short version
- Manufacturing load doesn't live at a headquarters three states away. It lives at one plant, on one meter, controlled by whoever runs that specific building.
- The federal government's own manufacturing energy survey (MECS) proves the load exists at scale, but it's published only as national and subsector tables, never as a named, callable list of facilities.
- ZoomInfo, D&B Hoovers, and a NAICS-filtered purchase list all roll up to the company level, so a 40,000-square-foot blending operation and an 800,000-square-foot continuous-process plant look identical on the list.
- The fix: use building square footage and headcount as load proxies, process type as an energy-intensity signal, and territory matched to your deregulated states, then reach the facilities manager who actually controls that address.
If you sell C&I energy supply, industrial fuel, demand response, or commercial solar into manufacturing plants, we've written before about the HQ-record mismatch that plagues territory sellers generally. For an energy seller it isn't an inconvenience, it's the whole basis of the deal: you can't size a load, an array, or a fuel contract off a company record with no building attached to it.
Why this matters right now
EIA expects US electricity sales to hit 4,135 billion kilowatthours in 2026, climbing again in 2027, and the agency names rising manufacturing activity as one of the drivers alongside data center growth. That's the federal government's own forecast saying industrial demand is climbing this year, not a vendor's pitch, and it's a large pool of load sitting behind addresses that current prospecting tools simply can't produce.
What's actually in a facility record
A useful prospecting list for this persona needs two things a company database has never had: a real building tied to a real address, and a way to estimate what's running inside it without a utility bill.
Facilities Finder tracks more than 600,000 plants, warehouses, and branches across all 50 states, each one geocoded and linked to its parent company. It ties 25 million employees to the specific facility they work at, not just a company name, and carries 20+ structured fields on every record. The two fields that matter most for sizing load are building square footage measured from satellite imagery and headcount drawn from company sites, filings, and professional networks: the same rough proxies an energy engineer already reaches for when there's no meter read available, now attached to a specific address instead of a guess.
The underlying data comes from satellite imagery for building size and location, company websites for products and processes, EPA and state environmental filings, business registries and public records, and professional networks, with AI agents reading every source and resolving it onto one record per facility. Manufacturing alone accounts for roughly 182,000 of those facilities, industrial services another 168,000, and warehouse and distribution sites about 98,000: the three segments where C&I load, fuel spend, and roof space concentrate.
How to build a working list instead of guessing
- Filter by building square footage and estimated headcount as load and roof-size proxies. Square footage isn't self-reported from a company website here, it's measured off satellite imagery, so a 40,000-square-foot blending operation and an 800,000-square-foot continuous-process plant stop looking identical the moment you filter by size.
- Search by process type in plain English, not NAICS code alone. AI search over facility descriptions finds phrases like "metal casting" or "plastics injection molding" directly, separating a heavy-process plant from a light-assembly shop sharing the same six-digit code.
- Filter by state or radius matched to your deregulated markets. Eighteen states plus DC have deregulated retail electricity, and a territory search keeps the list inside where you can actually write the contract.
- Reach the plant facilities or operations manager directly, matched by domain to that specific address, not a corporate sustainability contact with no authority over the roof or the rate schedule.
Pull those filters together and a facility's profile shows the satellite-measured building size and the reported headcount side by side on one record: the same two numbers an engineer already eyeballs to size a load, just attached to an address with a name and a phone number behind it.
Building size and headcount as load proxies, process type as an energy-intensity signal, and a facility-level contact instead of a corporate one: that's what turns a national survey number into a call list. See Best B2B Databases With Territory Mapping, Ranked for how facility-level data compares across providers generally.