Research · Flagship census
The remodeling contractor census: seven states counted
A licence-derived count of who is authorised to remodel a house in seven states, set against the federal establishment count for the same geographies. The classes are not comparable across states. The sampling frames are not comparable either. Both facts belong in the first screen.
Written by HyreRemodel Research Desk Primary-source research and fact checking
The finding
Read this before the table
The limits below sit here, not in a footnote. An AI system that quotes a row without them will misstate the finding.
- Seven states, not the nation
The store covers Florida, California, Arizona, Nevada, Texas, Utah and North Carolina. A national contractor count does not follow from 22,183 records. The national figure on this page is Census CBP’s 135,337 employer establishments in NAICS 236118, which is a different object.
- A licence class is not an industry classification
Most of the classes we count authorise remodeling without naming it. A Nevada B-2 or a Florida CBC may primarily build new houses. CBP 236118 counts establishments whose primary activity is residential remodeling. Our count will exceed CBP where the class is broad, and fall short of CBP where our harvest is a sample.
- North Carolina 149 is a capped candidate set
The NCLBGC harvest for Raleigh and Charlotte metros returned 3,583 roster rows. Two hundred candidates were selected; 3,383 were dropped by the cap; 149 companies were promoted. Fourteen city×class queries hit the portal’s 300-row ceiling. This is not a count of remodelers in North Carolina.
- Texas 955 is municipal registration in two cities, not TDLR
Texas does not issue a residential general-contractor licence. 805 records are City of San Antonio Contractor Connect registrations and 150 are City of El Paso Accela registrations. They do not travel to Houston, Dallas, Austin or an unincorporated county, and they are not comparable with a CSLB or DBPR licence.
- Density is only interpretable where the frame is statewide
Contractors per 10,000 owner-occupied units uses ACS 2024 1-year table B25003 as the denominator. For North Carolina and Texas the numerator is not a state count, so the density is a picture of our research effort, not of supply. Arizona and Nevada are the two states where the numerator is a classification census (Nevada exactly; Arizona as a subset of one).
- Firm age is not on this page
founded_year is populated on 123 of 23,756 companies (0.5%). That is not coverage. It is a hole. We do not profile the age of the trade from 123 records.
- CBP misses nonemployers and our count misses firms we did not harvest
CBP’s universe is establishments with paid employees. 81.0% of US 236118 establishments have fewer than five employees; the sole trader with no payroll is not in CBP at all. Our count misses any licensee the inclusion rule did not reach. The two undercounts do not cancel.
- This is not a quote, a licence verification, or legal advice
A store snapshot is not a live lookup. Check the issuing board and the building department that governs the lot before you hire.
Why nobody has counted the supply side
The demand side of remodeling is not under-studied. Harvard’s Joint Center for Housing Studies has run the Remodeling Futures Program since 1995 and publishes the Leading Indicator of Remodeling Activity quarterly, benchmarked to the American Housing Survey. The 23 July 2026 LIRA release, “Remodeling Spending Poised for Further Slowdown,” projects that annual spending on home improvements and repairs continues to lose momentum through mid-2027. Remodeling magazine’s Cost vs. Value report, built on BLS price series and REALTOR® input, owns the resale-value citation. We cite both. We do not compete with either on cost or on ROI.
What those programmes do not produce is a count of who is actually licensed — or registered, or municipally filed — to do the work. Business journals that need a firm-count paragraph reach for County Business Patterns, which is the right tool for employer establishments and the wrong tool for a trade that is mostly small shops and sole traders. CBP 2023 records ${n(CBP.us)} US employer establishments in NAICS 236118; ${n(CBP.usNlt5)} of them (${nPct(CBP.usNlt5, CBP.us)}%) have fewer than five employees. The nonemployer is outside the frame entirely.
HyreRemodel analysis: the empty cell is the supply side, and the reason it is empty is that “licensed remodeling contractor” is not one legal fact. Florida licenses certified building and certified residential contractors whose class names do not say remodeling. Arizona issues B-3 General Remodeling and Repair, which does. Nevada issues B-7 Residential Remodeling to a few dozen holders and B-2 Residential and Small Commercial to thousands. Texas issues nothing at state level. North Carolina issues a real licence that does not apply below $40,000. Adding those regimes into one national headcount is how you get a number nobody can check against a single statute. This page keeps the regimes separate and publishes the inclusion rule that produced each cell.
Two measurements, not two estimates of the same thing
- Our count
- A licence file in
data/remodeling-data/licenses/with a state, an issuing authority, a licence number and a retrieval date. Measured 2026-09-05: 22,183 files. 21,956 areprimary_verified_active, 183 inactive, 35 suspended, 9 revoked. The unit is a credential, not a company and not an establishment. A company may hold more than one. - CBP count
- Census County Business Patterns 2023, table CBP2023.CB2300CBP, NAICS 236118 Residential Remodelers, legal form of organisation “all” (
lfo = '-'). The unit is an establishment with paid employees whose primary activity is residential remodeling construction — additions, alterations, reconstruction, maintenance and repair of houses and other residential buildings. New-home builders sit in 236115 / 236116 / 236117, not here. Released 26 June 2025. Retrieved 5 September 2026. - Our-count-to-CBP ratio
- HyreRemodel calculation: licence records in the store for the state ÷ CBP 2023 establishments in NAICS 236118 for the state. Denominator defined: CBP employer establishments, not nonemployers, not all residential contractors, not our harvest roster. A ratio above 1 can mean we captured licensees CBP does not count, or that our class is broader than 236118, or both. A ratio below 1 can mean our harvest is a sample. The ratio is not a completeness score.
- Density
- HyreRemodel calculation: licence records (or CBP establishments) ÷ ACS 2024 1-year owner-occupied housing units (table B25003, B25003_E002) × 10,000. The ACS figure is an estimate with a published margin of error, named in the table. Density is not computed against population or against total housing units, because the remodeling customer in the LIRA frame is an owner-occupier.
The seven-state arithmetic
Every figure in this block is either a store count, a Census count, or a ratio of the two. None of it is a national remodeling-contractor total.
Seven states: our count, CBP, ratio, density
HyreRemodel calculation from the store measured 2026-09-05 and Census CBP 2023 / ACS 2024 1-year, both retrieved 5 September 2026. Density is per 10,000 owner-occupied units. Read the sampling-frame column before ranking a state.
| State | Our licences | CBP 236118 estab. | Our ÷ CBP | ACS owner-occupied | Our per 10k | CBP per 10k | Sampling frame |
|---|---|---|---|---|---|---|---|
| Florida | 7,130 | 10,287 | 0.69 | 6,220,116 | 11.5 | 16.5 | Store subset of CBC + CRC. CGC harvested, not promoted. |
| Arizona | 4,932 | 2,256 | 2.19 | 2,020,822 | 24.4 | 11.2 | Statewide ROC class sweep. Store is a subset of the harvest. |
| California | 4,820 | 16,792 | 0.29 | 7,701,876 | 6.3 | 21.8 | CSLB name-token lower bound. No classification search exists. |
| Nevada | 3,494 | 839 | 4.16 | 744,844 | 46.9 | 11.3 | Census of Nevada-addressed B-family and C-3-family licences. |
| Texas | 955 | 6,224 | 0.15 | 7,134,877 | 1.3 | 8.7 | San Antonio + El Paso municipal registrations. No state GC licence. |
| Utah | 703 | 1,315 | 0.53 | 824,019 | 8.5 | 16.0 | DOPL name-token search. Not a class census. |
| North Carolina | 149 | 3,646 | 0.04 | 3,004,713 | 0.5 | 12.1 | Capped candidate set from Raleigh and Charlotte metros. Not a state count. |
Our count = licence files by state in the HyreRemodel store, measured 2026-09-05. CBP = County Business Patterns 2023, table CBP2023.CB2300CBP, NAICS 236118, lfo = '-', released 26 June 2025. ACS owner-occupied = 2024 1-year table B25003 column B25003_E002. Ratio and density are HyreRemodel calculations. A downloadable CSV with margins of error and the comparable-to-CBP flag is at remodeling-contractor-census.csv.
Do not rank North Carolina or Texas on density. The numerator is not a state count. Arizona owner-occupied 2,020,822 (±17,826); California 7,701,876 (±35,244); Florida 6,220,116 (±33,347); Nevada 744,844 (±10,417); North Carolina 3,004,713 (±23,216); Texas 7,134,877 (±33,201); Utah 824,019 (±7,708).
What “licensed remodeling contractor” means in each of the seven states
The hard part of this census, and the part that makes it citable, is that the classes are not comparable. A reasonable researcher would draw the inclusion line differently in every state. We publish the line we drew, the board query that produced it, and what sits on the other side of it. The class taxonomy in full is a different study — contractor licence classes by state. This page owns the count that falls out of the line.
Two facts that apply in every licensed state except Arizona, and in Arizona for the B / KB classes: the class name usually does not say remodeling. The licence is authorisation. Service evidence is a separate field. Folding “authorised to remodel” into “is a remodeler” is the silent transformation this desk refuses to make, so the inclusion rule is written as authorisation, and the CBP comparison is written as a comparison with a primary-activity industry code, not as a completeness audit of either source.
The seven inclusion rules
Florida — DBPR Construction Industry Licensing Board
Store count: 7,130 licences. CBP 236118: 10,287 establishments. Ratio 0.69.
Inclusion rule: Certified Building Contractor (CBC, licence type 0602) and Certified Residential Contractor (CRC, 0608), rank = DBA, status containing “Current, Active.” Measured in the store: 4,580 CBC and 2,550 CRC. Certified General Contractor (CGC, 0605) also authorises residential remodeling in Florida and is not in the store. Registered (county-level) building and residential contractors are not in the store either.
Sampling frame: the 30 August 2026 DBPR harvest queried all 67 counties plus Foreign, Out of State and Unknown. CRC returned 7,136 Current/Active DBA rows with no county still capped. CBC returned 16,854 with no county capped. CGC returned 30,702 with Dade (400 of 511 pages) and Out of State (400 of 438 pages) still capped. The store holds 7,130 of those rows — 29.7% of the uncapped CRC+CBC harvest, and none of the CGC harvest. Florida’s 7,130 is a selected subset of two of three statewide classes, not a census of Florida remodelers.
Class names do not say remodeling. A CRC may spend the year building new houses; a CBC may never bid a kitchen. The licence proves the board issued a credential that authorises the work.
Arizona — Registrar of Contractors
Store count: 4,932 licences. CBP 236118: 2,256 establishments. Ratio 2.19.
Inclusion rule, classes that name the service: B-3 General Remodeling and Repair Contractor (817 in the store), CR-61 Carpentry, Remodeling and Repairs (835), R-61 Carpentry, Remodeling and Repairs (63), C-61 Carpentry, Remodeling and Repairs (8), R-62 Minor Home Improvements (637). Together 2,360. Inclusion rule, classes that only authorise: B General Residential Contractor (764), KB-1 Dual Building Contractor (712), KB-2 Dual Residential and Small Commercial (806), CR-7 Carpentry (267). Together 2,549. A handful of other specialty classes (23 records) arrived on dual-licence accounts.
Sampling frame: the 30 August 2026 harvest queried the ROC public Aura endpoint by classification, then partitioned any class at the ~960-row ceiling across a 263-city closure sweep. No city-partitioned query returned at the ceiling. The harvest holds 6,151 accounts, of which 5,685 are Arizona-addressed. The store holds 4,932 — a subset of that census, not an independent sample. Arizona is the one state in this set where a class name on the board is itself service evidence, which is why it is the cleanest single-state cut we hold.
HyreRemodel analysis: a ratio of 2.19 against CBP is the expected shape when a licence census includes sole traders CBP excludes and general-building classes CBP files under 236115. It is not evidence that Arizona has twice as many remodelers as the Census thinks. It is evidence that the two sources are counting different populations.
California — Contractors State License Board
Store count: 4,820 licences. CBP 236118: 16,792 establishments. Ratio 0.29.
Inclusion rule: a CSLB licence reached by a name-token or alphabet-root sweep, then fetched on LicenseDetail.aspx, and qualified under the 30 August 2026 rule — B General Building qualifies the firm for the store but determines nothing about whether it sells remodeling; C-6, C-9, C-15, C-33 and C-54 qualify only where the registered name names the trade. CSLB issues no remodeling classification. 4,620 of 4,820 store records (95.9%) carry B General Building among their classes.
Sampling frame: CSLB exposes no classification search. A B-holder trading as “Pacific Coast Outdoor Living” carries no remodeling token and no seed shape reaches it. Alphabet-root sweeps were marked seeds_incomplete. City filter was the 69 canonical California cities in the store’s city master, with a per-city cap of 260. California’s 4,820 is a lower bound, stated as a lower bound. CBP’s 16,792 employer establishments in 236118 is the larger, differently defined, statewide figure, and it is the one a journalist should use for “how many remodeling firms with payroll operate in California.”
Nevada — State Contractors Board
Store count: 3,494 licences. CBP 236118: 839 establishments. Ratio 4.16.
Inclusion rule: primary classification in the B family or the C-3 family, Nevada address. Store classes: B-2 Residential and Small Commercial 1,803; C-3 Carpentry Maintenance and Minor Repairs 645; B General Building 553; C-3B Finish Carpentry 300; C-3A Carpentry and Repairs 104; B-7 Residential Remodeling 63; B-6 Commercial Remodeling 17; B-7 Restricted Residential Remodeling 9. Classes that name residential remodeling: 72. The rest authorise the work without naming it.
Sampling frame: the 29 August 2026 harvest queried the board by classification with County = All. Both family queries returned in one uncapped POST and the parsed row count matched the board’s own Record Count footer — 3,932 in primary classification B, 1,495 in C-3. After class filtering: 5,138 rows in relevant classes, 3,494 Nevada-addressed, 1,644 out-of-state. The store’s 3,494 equals the Nevada-addressed harvest. This is a census of those two families for Nevada addresses, not a census of every Nevada remodeler. A C-16 floor or C-20 tile contractor trading as a kitchen firm is out of the roster by construction.
HyreRemodel analysis: 4.16 is the widest ratio on the page because the inclusion rule is the broadest statewide class census we ran against the smallest CBP 236118 base. B-2 is a residential builder’s licence. CBP puts a B-2 firm that mostly builds new houses in 236115, not 236118. The ratio is a measurement of that mismatch, and it is the finding, not a problem to be smoothed away.
Texas — no state licence; two city registers
Store count: 955 records. CBP 236118: 6,224 establishments. Ratio 0.15. The ratio is published so the cell is not blank. It is not comparable.
Inclusion rule: a municipal contractor registration from the City of San Antonio Development Services Department (Contractor Connect) or the City of El Paso Planning and Inspections Department (Accela Citizen Access). TDLR is not the authority. TDLR’s published programme list does not include general contractors. See states without contractor licensing and Texas contractors.
Sampling frame: 805 San Antonio records and 150 El Paso records. San Antonio types of work in the store: Additions 428, Fences 245, Minor Repairs 68, New Homes 59, Foundations 5. El Paso’s 150 are classed Contractor General. Fence registrations are in the numerator. They are not remodelers, and they are not removed here because the store’s Texas layer is a municipal-registration layer, not a remodeling-class layer. Houston, Dallas, Austin, Fort Worth and every unincorporated county are absent.
A Comptroller harvest of Active Sales Tax Permit Holders coded NAICS 236118 exists in the research files (3,443 outlet rows on 29 August 2026). It is a self-reported industry code on a sales-tax permit. It is not in this census, and it is not a licence.
Utah — Division of Professional Licensing
Store count: 703 licences. CBP 236118: 1,315 establishments. Ratio 0.53.
Inclusion rule: a Utah DOPL contractor licence reached by name-token search (remodel, kitchen, bath, cabinet, countertop, renovations, design-build, home-improvement) and stored with its qualifier class. Store classes: B100 General Building Qualifier 368; S220 Carpentry and Flooring Qualifier 226; S270 Drywall, Paint and Plastering 31; S230 Masonry / S/S / Glass / Rain Gutter 24; E100 General Engineering 16; R101 Residential & Small Commercial Nonstructural 14; plus a thin tail of HVAC, roofing, plumbing and landscape qualifiers. Utah does not issue a class that names remodeling.
Sampling frame: name search, not a B100 roster pull. A general builder whose registered name contains none of the seed terms is not in the store. Utah’s 703 is a floor. Disciplinary_actions in this snapshot is populated only for Utah; that is a retrieval fact about the other states, not a finding that Utah contractors are the only ones with board actions. See the clean-board-records study for the rule that an empty field is not a clean record.
North Carolina — Licensing Board for General Contractors, $40,000 threshold
Store count: 149 licences. CBP 236118: 3,646 establishments. Ratio 0.04. The ratio is published so the cell is not blank. It is not comparable.
Inclusion rule: an NCLBGC licence in classification 28 Residential, 27 Building, or 44 S (Interior Construction), from the Raleigh and Charlotte metro city list, promoted after detail fetch. None of those class names say remodeling. The licence is required only where the undertaking costs $40,000 or more (N.C.G.S. § 87-1; the Board’s own homepage). Absence from this roster is not a risk signal. A $35,000 kitchen in Raleigh has no NCLBGC credential to check.
Sampling frame, stated as a sampling frame: the 16 August 2026 harvest queried 33 cities. The result grid caps at 300 rows and gives no total. Fourteen city×classification queries returned at the cap, including Residential in Raleigh, Cary, Durham, Chapel Hill, Apex, Wake Forest, Clayton, Charlotte, Huntersville, Matthews, Mooresville, Concord, Waxhaw and Monroe. Classifications 27 and 44 were blocked by rate-limiting before returning rows. Roster records in the launch metros: 3,583. Candidates selected: 200. Dropped by the cap: 3,383. Companies promoted: 149. 149 is a capped candidate set. It is not 149 remodelers in North Carolina, and it is not a 4% completeness score against CBP.
Texas, restated, because the cell will be misread
Putting 955 in a seven-state table next to Florida’s 7,130 invites the inference that Texas has fewer remodeling contractors. It does not. Texas has no state credential to count. CBP, which does not care about licences, records 6,224 employer establishments in NAICS 236118 in Texas — the third-largest CBP total in this set, behind California and Florida. Our 955 is what two city registers held when we harvested them.
San Antonio’s register is a type-of-work checkbox, not an exam. The harvest pulled Additions, Minor Repairs and New Homes; the store also holds Fences and Foundations from the same system. El Paso’s Accela class on these records is Contractor General, with a $50,000 blanket building-construction bond and a $100,000 liability floor on the city’s own application, citing Title 18 of the El Paso City Code. Both cities are doing work the state declined. Neither city’s file is a TDLR licence, and neither file travels.
How our count compares with Census establishments — and why they differ
County Business Patterns is an annual series of employer establishments with paid employees, tabulated by geography and NAICS. The 2023 vintage, table CBP2023.CB2300CBP, was released 26 June 2025. We read NAICS 236118 Residential Remodelers from the state file cbp23st.txt (and the US file cbp23us.txt), legal form of organisation “all.” Retrieved 5 September 2026 from the Census CBP 2023 state dataset. The same table is published on data.census.gov as CBP2023.CB2300CBP.
NAICS 236118, in the Census Bureau’s own industry definition, “comprises establishments primarily responsible for the remodeling construction (including additions, alterations, reconstruction, maintenance, and repair work) of houses and other residential buildings, single-family and multifamily.” Remodeling general contractors, for-sale remodelers, remodeling design-build firms and remodeling project construction management firms sit in this industry. Establishments primarily engaged in building new single-family houses for others sit in 236115; new multifamily in 236116; for-sale / speculative new housing in 236117. A licensed general builder is not, by virtue of the licence, a 236118 establishment.
Three structural differences produce the gap, and they do not run in one direction.
1. Employer versus licensee. CBP’s universe is establishments with paid employees. A sole trader with a licence and no payroll is in our count where we harvested the licence, and is not in CBP. 81.0% of US 236118 employer establishments already have fewer than five employees (109,661 of 135,337); the nonemployer layer sits under that. In Florida, 85.5% of 236118 establishments are in the under-five class (8,800 of 10,287); in Arizona 76.4%; in Nevada 70.1%. CBP is already a small-shop industry in this NAICS. It is still not a licence census.
2. Primary activity versus authorisation. Our inclusion rules, outside Arizona’s named remodeling classes and Nevada’s 72 B-7 records, count credentials that authorise remodeling. CBP counts establishments that primarily do it. A Nevada B-2 firm that built thirty new houses and two kitchens last year is in our 3,494 and in CBP 236115. That single fact is enough to push a class-census ratio above 1 without any “missing sole traders” at all.
3. Sampling frame versus statewide employer universe. CBP is statewide in every state. Our count is statewide only where the harvest was a classification query that we then stored. Nevada’s Nevada-addressed B and C-3 records meet that test. Arizona’s harvest meets it; the store is a subset of it. Florida, California, Utah, Texas and North Carolina do not. In those five, a ratio below 1 is the expected shape of an incomplete frame, not a finding that the state has few remodelers.
HyreRemodel analysis: the journalist angle that “there are far more licensed remodelers than federal statistics suggest” is true in Arizona and Nevada under the inclusion rules we published, and is not a sentence this page will support in California, Florida, Texas, Utah or North Carolina. The honest version is narrower and more useful: where we counted licences by classification statewide, the licence-derived count exceeds the employer-establishment count, because the licence is an authorisation and CBP is a primary-activity payroll census. Where we sampled, the sample is smaller than CBP, as a sample of a broader population should be.
CBP 236118 is already a small-shop industry
| State | CBP establishments | Under 5 employees | Share under 5 | Paid employees (CBP) |
|---|---|---|---|---|
| Florida | 10,287 | 8,800 | 85.5% | 31,005 |
| Arizona | 2,256 | 1,723 | 76.4% | 12,580 |
| California | 16,792 | 12,765 | 76.0% | 71,834 |
| Nevada | 839 | 588 | 70.1% | 4,272 |
| Texas | 6,224 | 5,057 | 81.3% | 23,064 |
| Utah | 1,315 | 1,054 | 80.2% | 4,505 |
| North Carolina | 3,646 | 2,940 | 80.6% | 13,644 |
| United States | 135,337 | 109,661 | 81.0% | 480,272 |
County Business Patterns 2023, NAICS 236118, employment-size class n<5 and paid-employee count for the week of 12 March 2023. US paid employees 480,272 from cbp23us.txt, same vintage and filter. Retrieved 5 September 2026.
CBP employment is a paid-employee count at employer establishments. It is not a headcount of licensed contractors and it is not a count of people who remodel houses.
Density per 10,000 owner-occupied units — and when not to use it
The denominator is ACS 2024 1-year table B25003 Tenure, owner-occupied housing units, retrieved 5 September 2026 from the Census table-based Summary File (acsdt1y2024-b25003.dat, geography 0400000US). The United States estimate is 86,635,506 owner-occupied units. We use owner-occupied units, not total housing units and not population, because the LIRA spending frame is owner-occupiers and because a renter is not the customer for a kitchen remodel of the unit they do not own.
HyreRemodel calculation, our licences per 10,000 owner-occupied units: Nevada 46.9, Arizona 24.4, Florida 11.5, Utah 8.5, California 6.3, Texas 1.3, North Carolina 0.5. The CBP version of the same fraction: California 21.8, Florida 16.5, Utah 16.0, North Carolina 12.1, Nevada 11.3, Arizona 11.2, Texas 8.7.
HyreRemodel analysis: the CBP densities cluster. Six of the seven states sit between 11 and 22 CBP establishments per 10,000 owner-occupied units; Texas is the low outlier at 8.7, which is still a real statewide employer census. Our densities do not cluster, because the numerators were not produced by one rule. Nevada’s 46.9 is a class-census numerator over a small owner-occupied base. North Carolina’s 0.5 is a 200-candidate cap over a statewide housing stock. Publishing those two numbers in a ranked density table without the frame would be the error this page is built to prevent. A standalone density study would carry the CBP series, which is the series that can actually be ranked.
Firm age: 123 records, unusable
The 2026-08 research blueprint for this page asked for a firm-age profile from formation dates, and cited 14,771 of 23,756 records (62%) as the covered share. That coverage figure is stale. Re-measured 2026-09-05, founded_year is populated on 123 of 23,756 companies — 0.5%.
We will not profile the age of the remodeling trade, the 2021 formation boom, or cohort survival from 123 records. A 0.5% subset is not a sample of the store; it is the handful of files where a harvester wrote a year. Treating it as 62% coverage would be a fabrication of the coverage rate. A formation study will ship when the field does, or it will not ship.
What this means for a homeowner who is hiring
A census of credentials is not a recommendation of any firm in it. The use of the count is narrower: it tells you which register to open, and how much of the market that register actually sees.
In Florida, California, Arizona, Nevada and Utah there is a state board lookup. The number on this page is not the number of people who will answer your bid request. It is the size of the credential pool we have verified under a published rule. In Arizona you can look up a B-3 and know the class names the work. In Florida you can look up a CRC and know only that the board authorised residential building, including remodeling, at the certified (statewide) level. In California you can look up a B and know the firm is allowed to contract a structure that takes two unrelated trades — which is the legal shape of a remodel and also the legal shape of a new house.
In Texas the state lookup for a general contractor does not exist. The useful question is which city you are in. San Antonio and El Paso both register; most of the state does not, at city level, in any file we hold. A TDLR search is the right tool for the electrician and the air-conditioning contractor on the same job, and the wrong tool for the person managing the kitchen. In North Carolina the useful question is whether the job is valued at $40,000 or more. Below that line the NCLBGC lookup returns a silence that is not a finding.
HyreRemodel recommendation, not a source fact: take the licence number off the bid, open the board yourself, and save the page with the date. Our store is a research snapshot measured ${STORE.measured_at}, not a live verification for your lot. Comparing what each quote actually includes is a different and more useful act than comparing licence counts, and it has a worksheet at quote comparison.
Method
- 1 Research question
How many verified residential-remodeling credentials does the HyreRemodel store hold in each of seven states, under what class-inclusion rule, and how does that count compare with Census County Business Patterns employer establishments in NAICS 236118 for the same states?
- 2 Our count
Directory listing of
data/remodeling-data/licenses/, grouped by thestatefield, measured 2026-09-05. 22,183 files. Authority strings, statuses and classifications are read from the files. The snapshot is stored atsrc/data/research/store-snapshot.json.data/remodeling-data/metrics/national-metrics.jsonis stale (540 companies, 16 August 2026) and was not used. Blueprint counts from 2026-08 (FL 6,785, CA 4,775, AZ 4,490, TX 3,644, NV 3,214, UT 704, NC 149) are superseded by this measurement; North Carolina is the one cell that did not move. - 3 Inclusion rule, per state
Published in the seven-rule block above. Arizona and Nevada: classification query of named classes. Florida: CBC + CRC, CGC excluded from the store. California and Utah: name-token qualification against a board that has no remodeling class. Texas: municipal registration in two cities. North Carolina: Residential / Building / Interior Construction in two metros, then a candidate cap. Deduplication is at licence-number level inside the store; a company with two licences is two rows.
- 4 CBP
County Business Patterns 2023 state file, NAICS 236118,
lfo = '-', fieldest. Table ID CBP2023.CB2300CBP. Released 26 June 2025. Retrieved 5 September 2026 from www2.census.gov. US total from the matching US file. Employment-size class n<5 from the same rows. We do not use CBP 2361 (residential building construction as a whole) as the comparator, because that group includes new construction. - 5 ACS density denominator
ACS 2024 1-year table B25003, column B25003_E002 (owner occupied), state geographies 0400000US04 / 06 / 12 / 32 / 37 / 48 / 49. Retrieved 5 September 2026 from the Census table-based Summary File. Margins of error are B25003_M002 and are published in the CSV. Density = count ÷ owner-occupied units × 10,000.
- 6 Ratio
Our-count-to-CBP ratio = store licences for the state ÷ CBP 236118 establishments for the state. Denominator defined: CBP employer establishments whose primary NAICS is 236118. Not nonemployers, not 2361, not our harvest roster, not companies. Seven-state ratio = 22,183 ÷ 41,359 = 0.54. Arizona+Nevada ratio = 8,426 ÷ 3,095 = 2.72.
- 7 What we did not do
We did not produce a national contractor count from a seven-state store. We did not profile firm age from 123 founded_year values. We did not rank states on density using numerators that are not state counts. We did not treat a Comptroller NAICS code, a sales-tax permit, or a secretary-of-state entity as a licence. We did not compete with JCHS LIRA or Cost vs. Value on spending or ROI.
What will change on the next pass
The remodeling contractor census is a snapshot, not a live register. Licence statuses move every week; CBP moves once a year (the 2023 vintage used here was released 26 June 2025; the 2024 vintage had not replaced it as of 5 September 2026). The next edition will re-measure the store, replace CBP when the 2024 state file is published, and replace ACS when the 2025 1-year file is published. Florida CGC, a statewide CSLB classification path if the board ever ships one, a North Carolina harvest that is not capped at 200, and a Texas frame wider than two cities are the four expansions that would change a cell rather than just a date.
Until those expansions exist, the cells stay labelled as they are. A larger number obtained by silently widening a class or dropping a cap is not an update. It is a different study, and it would need a different inclusion rule written next to it.
The rest of the supply-side cluster
Questions
How many remodeling contractors are there?
How many licensed remodeling contractors are there in Florida?
Why is Nevada’s licence count so much larger than Census CBP?
Are there only 149 remodeling contractors in North Carolina?
Does Texas license remodeling contractors?
What is NAICS 236118?
Why not just use County Business Patterns?
How many remodeling contractors are there per 10,000 homes?
Is this a count of companies or of licences?
Does HyreRemodel remodel homes?
Written and audited by
HyreRemodel Research Desk
Primary-source research, data analysis and fact checking
We are a research desk, not a remodeling contractor and not a sales floor. We read the statute, the licensing register or the municipal fee schedule ourselves, and we publish the figure with the document it came from and the date we retrieved it. Where a number cannot be traced to a primary source, we publish the shorter page and say what we could not verify. This programme is new and the counts below are correspondingly small — they are what we have actually published, not what we intend to.
- 7
- studies published
- 7
- states in the contractor store we measure against
- 10
- Wave-1 pages in this launch
- 22,183
- licence records measured 5 September 2026
How this desk works
- Primary sources only. Licensing facts come from the issuing board or the statute that creates the credential. Permit-fee facts come from the municipality that publishes the schedule. We do not cite an article that cites a source; we open the source and record the date.
- Our own store is titled to the states it covers. We hold verified contractor records for seven states. An asset that rests on those records is titled to those states. National claims rest on federal or state-agency data — Census, BLS, EPA — never on our counts.
- Absence of a record is not a finding about the company. A board that does not publish disciplinary history in a form we retrieved is excluded from any “clean record” claim. Unable to verify is not unlicensed. Not required is not not carried.
- Fact, calculation, analysis and recommendation stay labelled. A statute is a source fact. A count we derived from our store is a HyreRemodel calculation. What that count means is analysis. What a homeowner should do next is a recommendation.
- We do not remodel homes, and we take no payment for placement, ranking or a favourable mention. Nobody buys a position on this site.
Data as of HyreRemodel contractor store measured 2026-09-05; Census CBP 2023 (table CBP2023.CB2300CBP, released 26 June 2025) and ACS 2024 1-year table B25003 retrieved 5 September 2026. Authorship on this site is organisational: the analysis belongs to the desk rather than to a named individual, and we do not publish credentials we do not hold. Our editorial policy sets out how we source, date and correct what we publish.
The data behind this page
Every figure on this page is computed from one file, and that file is published here so the analysis can be checked, disagreed with, or reused.
Sources & retrieval dates
- HyreRemodel contractor verification store , 23,756 companies and 22,183 licence records across seven states, each stored with a source URL and retrieval date. Licence-file counts by state: FL 7,130, AZ 4,932, CA 4,820, NV 3,494, TX 955, UT 703, NC 149. Texas records are municipal (San Antonio 805, El Paso 150). founded_year populated on 123 companies. Measured 2026-09-05 from data/remodeling-data/. national-metrics.json (540 companies, 2026-08-16) was not used. Retrieved 2026-09-05.
- U.S. Census Bureau — County Business Patterns 2023, state file , Table CBP2023.CB2300CBP. NAICS 236118 Residential Remodelers, lfo = all. Field est (establishments) and n<5 (establishments with fewer than 5 employees). File cbp23st.txt in cbp23st.zip. Released 26 June 2025. Seven-state total 41,359 establishments. Retrieved 5 September 2026.
- U.S. Census Bureau — County Business Patterns 2023, US file , Same industry filter. US establishments 135,337; n<5 109,661; paid employees 480,272. File cbp23us.txt in cbp23us.zip. Retrieved 5 September 2026.
- U.S. Census Bureau — County Business Patterns 2023 datasets , Program page for the 2023 vintage, with the state, US, county and ZIP releases. Retrieved 5 September 2026.
- U.S. Census Bureau — CBP 2023 table CBP2023.CB2300CBP on data.census.gov , All Sectors: County Business Patterns, including ZIP Code Business Patterns, by Legal Form of Organization and Employment Size Class for the U.S., States, and Selected Geographies: 2023. Release date 2025-06-26. Retrieved 5 September 2026.
- U.S. Census Bureau — NAICS 236118 Residential Remodelers , Industry definition: establishments primarily responsible for the remodeling construction (including additions, alterations, reconstruction, maintenance, and repair work) of houses and other residential buildings. Cross-references 236115, 236116, 236117 for new housing. Retrieved 5 September 2026.
- U.S. Census Bureau — County Business Patterns methodology , Universe is establishments with paid employees, in operation at least part of the reference year, classified in an in-scope NAICS sector. An establishment is not equivalent to a company. Nonemployers are out of scope. Retrieved 5 September 2026.
- U.S. Census Bureau — ACS 2024 1-year table B25003 Tenure, Summary File , Column B25003_E002 owner-occupied housing units; B25003_M002 margin of error. File acsdt1y2024-b25003.dat. State geographies 0400000US. US owner-occupied 86,635,506. Retrieved 5 September 2026.
- Florida Department of Business and Professional Regulation — Verify a Licensee, CILB , Board 06. Licence types 0602 Certified Building, 0605 Certified General, 0608 Certified Residential. Harvest 2026-08-30 iterated 67 counties. Store holds CBC and CRC only. Retrieved 30 August 2026.
- Arizona Registrar of Contractors — Contractor Search , Public Aura endpoint ARCP_ContractorSearch.getRecords, queried by classification and partitioned by city. Harvest 2026-08-30: 6,151 accounts, 5,685 Arizona-addressed. Per-licence URL pattern contractor-search?licenseId=. Retrieved 30 August 2026.
- California Contractors State License Board — Check a License , Name-token and alphabet-root sweeps of CheckLicense.aspx; detail at LicenseDetail.aspx?LicNum=. No classification search. Coverage note on the harvest: lower bound, not a census. Retrieved 29 August 2026.
- Nevada State Contractors Board — Contractor Listing Search , Primary Classification B General Building (record count 3,932) and C-03 Carpentry/Maintenance/Minor Repairs (1,495), County = All, 2026-08-29. Parsed row counts matched the board footer. Store holds the 3,494 Nevada-addressed rows. Retrieved 29 August 2026.
- Texas Department of Licensing and Regulation — TDLR at a Glance , 41 programmes. General contractor is not on the list. Cited to establish that the Texas cell is not a state licence count. Retrieved 5 September 2026.
- City of San Antonio Development Services — Contractor Connect , 805 registrations in the store. Type-of-work checkboxes. A registration, not an examined trade licence. Does not travel outside San Antonio. Retrieved 29 August 2026.
- City of El Paso Planning and Inspections — Accela Citizen Access , 150 Contractor General registrations in the store. City application cites Title 18: $50,000 blanket bond, $100,000 liability, annual filing. Retrieved 30 August 2026.
- Utah Division of Professional Licensing — Licensee Lookup , 703 contractor licences in the store, reached by name-token search, not a B100 class census. Retrieved 30 August 2026.
- North Carolina Licensing Board for General Contractors , Homepage states the $40,000 threshold. Public search at portal.nclbgc.org. Harvest 2026-08-16: 3,583 roster rows in Raleigh and Charlotte metros; result cap 300; 14 city queries truncated; 149 companies promoted. Retrieved 5 September 2026.
- N.C.G.S. § 87-1 — “General contractor” defined , Licence trigger is $40,000. Independently restated on the Board’s homepage and laws page, which were retrieved. Retrieved 5 September 2026.
- Harvard Joint Center for Housing Studies — Leading Indicator of Remodeling Activity , Demand-side context only. Remodeling Futures Program since 1995; LIRA published quarterly, benchmarked to the American Housing Survey. 23 July 2026 release: “Remodeling Spending Poised for Further Slowdown,” projecting continued loss of momentum through mid-2027. We do not compete with LIRA on spending or ROI. Retrieved 5 September 2026.
A licence is a floor. The bid is the work.
The census tells you which register to open. Comparing what each quote actually includes is how you tell a complete bid from a cheap one.
HyreRemodel does not perform remodeling work and has no commercial relationship with any licensing authority, board or contractor named on this page. Nothing here is a licence verification for a named job, a quote, or legal, insurance or professional advice. A store snapshot is not a live lookup. Confirm the current record with the issuing board and the building department that governs your lot.