PermitExport
July 19, 2026

One Year, Four Metros: What 221,175 Building Permits Reveal About Construction Demand

A trailing year of building permits across Austin, San Antonio, Colorado Springs and Sacramento — 221,175 records. What the data shows about HVAC, roofing, solar and EV demand, and why some metros can answer questions others can't.

One Year, Four Metros: What 221,175 Building Permits Reveal About Construction Demand

The numbers. Over a trailing 12-month window (mid-2025 → mid-2026) we compiled 221,175 building-permit records across four US metros — Austin–Round Rock, San Antonio–New Braunfels, Colorado Springs, and Sacramento–Roseville. The counts below are what each metro’s own permit classifications report, stated as counts.

  • HVAC is the quiet giant of residential permitting: 10,810 HVAC/mechanical permits in Austin, 12,158 in San Antonio–NB, 7,469 in Sacramento — the only vertical strong enough to report in three of the four metros.
  • San Antonio–NB recorded 5,981 roofing permits in twelve months — a dated, official-record measure of roofing demand no survey can match. Sacramento adds 4,537.
  • Sacramento is the only studied metro where solar and EV-charger activity are directly readable: 1,986 solar permits and 660 EV-charger permits in one year.
  • Our files capture 80–89% of officially reported new-residential activity where a like-for-like comparison is computable: San Antonio 80.7%, Colorado Springs 89.2%, Sacramento ≥80.9% — validated month-by-month against the US Census Building Permits Survey.

There is no national permit feed

US building-permit data lives in hundreds of separate city and county systems, and at scale it is close to unusable as-is. Anyone who wants to answer a simple question — how many roofs were replaced in this metro last year? — has to go jurisdiction by jurisdiction and reconcile four different ways of describing the same work. We did that once, so the answer is a query instead of a project.

Four metros, four ways of describing a building

MetroRecords (12 mo)What its data can tell you
Austin–Round Rock, TX60,450Trade-level detail (Electrical 15,827; Plumbing 14,609; Mechanical 10,631) with clean work classes — but no roofing or solar vocabulary
San Antonio–New Braunfels, TX87,829 line items (~75,100 distinct permits)The most granular permit-type vocabulary studied
Colorado Springs, CO20,994Numeric construction-type codes rather than words. Near-100% valuation reporting ($3.27B declared, median $15,000)
Sacramento–Roseville, CA51,902The most descriptive work classifications — vertical signals readable directly

The finding that matters: the same question is answerable in one metro and invisible in another. “How many re-roofs?” is a one-line query in San Antonio and Sacramento, and unanswerable from Austin’s trade-permit vocabulary alone. A national number assembled by naive keyword-matching would quietly undercount some metros and overcount others.

The vertical signals

Twelve-month permit counts, shown only where the metro’s own permit classifications name the vertical directly:

VerticalAustinSan Antonio–NBColorado SpringsSacramento
HVAC / mechanical10,81012,158n/r¹7,469
Roofing / re-roofn/r¹5,981n/r¹4,537
Solar / PVn/r¹~871²n/r¹1,986
Pool / span/r¹~427²n/r¹738
EV chargern/r¹n/r¹n/r¹660

¹ Not reportable from that metro’s permit classifications — publishing it would mean publishing a false zero or an unverifiable guess. ² Detected, but with attribution caveats — treat as indicative rather than exact.

And one we refuse to publish anywhere: kitchen & bath remodels. Every studied jurisdiction bundles them into generic alteration/remodel classes. In Sacramento, counting the bundle would overstate the true kitchen-and-bath number roughly 34-fold. Vendors selling “kitchen remodel leads from permit data” are, in these four metros at least, selling the bundle.

How complete is it? We checked against the Census

Where a like-for-like comparison is computable — new single-family permits against the US Census Building Permits Survey’s 1-unit authorizations, month by month over 11 published months — our files capture:

  • San Antonio–NB: 80.7% of officially reported new-residential volume
  • Colorado Springs: 89.2%
  • Sacramento: ≥80.9% (a floor)
  • Austin: not computable as like-for-like — reported as N/A rather than estimated

The Census survey counts only new housing units. Our files carry 3.5×–15.5× more records than those aggregates — trade permits, remodels, re-roofs, equipment change-outs — which is exactly where the vertical signals above live.

What this means if you buy permit data

  1. Official records beat surveys for demand measurement. A re-roof permit is a household that actually spent money, dated and geolocated — not a stated intention.
  2. What a metro can tell you varies enormously. Sacramento is a vertical-intelligence source; Colorado Springs is a valuation and new-construction source.
  3. Valuation reporting is wildly uneven — near-100% in Colorado Springs and 94% in Sacramento, against 8% in Austin and 2% in San Antonio. Cross-metro dollar comparisons drawn from raw permit data are not honest without saying so.
  4. Ask any vendor what their numbers exclude. If they can’t tell you which verticals their data genuinely supports and which it doesn’t, the number is a guess.

About

PermitExport sells historical building-permit record files for US metros, compiled fresh on order. Independent compilation — not an official record.

Get a data file → · Questions? Contact us →