The Diversified Support Services Data Guide (2026)

Diversified Support Services — administrative and support and waste management, NAICS 56 — is one of the rare verticals where the public statistical system covers nearly everything. As of August 2026, Datadory catalogs 21 datasets for this industry: 16 primary records and 5 related records drawn from neighboring industries such as automotive retail, data processing, electrical components and leisure facilities.

US federal statistics anchor the slice. BLS contributes the Quarterly Census of Employment and Wages (QCEW) — quarterly employment, establishment counts and wages by six-digit NAICS and county from 2004 through Q4 2025 (~815,000 rows per recent-year file) — and OEWS annual wage estimates for roughly 830 occupations including janitors, security guards and maintenance workers. The Census Bureau adds the Quarterly Services Survey of NAICS 56 revenue since Q4 2003 and Statistics of U.S. Businesses (SUSB) firm-size counts down to county level.

International coverage comes from two UK registers and Canada. Statistics Canada publishes NAICS 56 payroll tables — 17,598 series in table 14-10-0201 alone — ONS BRES breaks the UK's 2.73 million enterprises down by SIC class, and Companies House makes company-level registrations filterable to private security, facilities support and cleaning.

US labor and market sizing

Quarterly Census of Employment and Wages (QCEW) — NAICS 56 (quality score 9) is the near-census backbone: quarterly employment, establishment counts and wages by six-digit NAICS industry and geography at NAICS levels 0 through 6, running 2004 through fourth quarter 2025 including annual aggregates. Recent-year files hold about 815,000 rows — six files totalling roughly 680 MB cover the full 2004–Q4 2025 span. BLS suppresses small county-industry cells that would disclose cooperating employers.

Occupational Employment and Wage Statistics (OEWS) — support occupations (quality score 8) benchmarks pay for roughly 830 occupations including janitors, security guards and maintenance workers, by metropolitan statistical area, state and industry, with hourly and annual wage variants. The catalog record spans 2009 through 2026 first quarter as a single table of about 580,000 rows. Treat the estimates as a snapshot in time rather than a chainable series.

Statistics of U.S. Businesses (SUSB) (quality score 9) counts firms, establishments, employment, payroll and receipts by six-digit NAICS and enterprise employment-size class at US, state, MSA, county (from 2000) and congressional-district (from 2015) level. The 2022 US/state six-digit dataset alone runs about 56 MB of CSV, spanning ~6.4 million firms and ~8.3 million establishments across all industries. Values carry noise injection and disclosure suppression, so cells flagged G/H/J/S should be treated as approximations.

Quarterly Services Survey (QSS) (quality score 7) reports national NAICS 56 revenue and expenses each quarter — covered since Q1 2009, with the benchmark file reaching back to Q4 2003. It reports United States totals only, with no state or metro breakdown.

Safety, contracting and corporate filings

SAM.gov Entity Registry (quality score 8) lists every entity registered to do business with the US federal government, searchable by UEI, CAGE code, registration status and NAICS/PSC codes such as janitorial, security services, staffing and facilities management. Extracts cap at 1,000,000 records per request, with support-services NAICS subsets in the tens of thousands. Records containing D&B Open Data carry third-party restrictions from Dun & Bradstreet.

SEC EDGAR Full-Text Search API (quality score 9) indexes tens of millions of filing documents since January 2001, returning CIK, company name, SIC code, filing date and accession number per hit. A query for 'staffing services' restricted to 10-K forms matched 2,486 filings.

Wikipedia List of S&P 500 Companies (GICS sub-industry) (quality score 6) rounds out the US set with a community-maintained table of constituents — 503 rows covering 500 companies due to dual share classes — carrying ticker, GICS sub-industry, headquarters, SEC CIK and founding year, refreshed daily by editors.

UK and Canadian registers

UK Business Register and Employment Survey (BRES) (quality score 9) snapshots every VAT and/or PAYE registered UK business — 2.73 million enterprises and 3.2 million local units — by UK SIC 2007 class (including SIC 80–82 support services), region, district, constituency, employment size band and legal status. The current edition is a March 2025 IDBR snapshot published 24 September 2025; figures are disclosure-controlled and rounded to base 5.

UK Companies House — SIC 80–82 company registrations (quality score 8) provides company-level status, incorporation dates and registered offices filterable by SIC code: 44,686 companies carry SIC 80100 (private security), 20,302 have 81100 (facilities support) and 63,603 carry 81210 (building cleaning), including dissolved firms. The full register covers ~5 million live companies split into 7 parts.

Statistics Canada — NAICS 56 data tables (quality score 9) deliver monthly payroll employment, hours and earnings by industry and geography for all 14 provinces and territories. Table 14-10-0201 alone carries 17,598 series and ~5.3 million datapoints from January 2001 through May 2026, with bulk archives measuring about 128 MB.

Each major support-services workflow maps to specific named sources rather than generic categories:

1. Size local janitorial, security and staffing markets before expansion. Pull quarterly employment, establishment counts and wages for NAICS 56 by county, metro and state from BLS QCEW — 2004 through Q4 2025, about 815,000 rows per recent-year file — then segment firm targets with SUSB's enterprise employment-size classes down to county level.

1. Benchmark pay for frontline occupations. Compare janitor, security guard and maintenance worker wage estimates by metro area in OEWS, which covers roughly 830 occupations across California MSAs and OES survey regions, with equivalent national files from BLS.

1. Screen vendors' safety records before awarding facility-services contracts. Check establishment-level injury and illness filings (about 383,000 rows in the 2025 300A summary), citations, penalties and severe injury reports in OSHA Enforcement & Injury Data, refreshed daily.

1. Build prospect lists of federal contractors. Filter the SAM.gov Entity Registry by UEI, CAGE code and NAICS/PSC codes for janitorial, security services, staffing and facilities management — hundreds of thousands of registered entities, extractable up to 1,000,000 records per request.

1. Track revenue momentum in administrative support and waste management. Follow the Quarterly Services Survey's national NAICS 56 revenue and expense estimates back to Q4 2003, usable directly or alongside GDP work as an economic indicator.

1. Analyze publicly traded staffing and facility-services peers. Query SEC EDGAR full-text search across filings since 2001 — 'staffing services' in 10-Ks alone matched 2,486 filings, each hit returning CIK, SIC code and accession number.

1. Map the UK business population by service line. Break ONS BRES's 2.73 million enterprises down by SIC 80–82 class, region and constituency, then join company-level status and incorporation dates from Companies House for SIC 80100/80200, 81100 and 81210.

1. Benchmark Canadian payroll employment province by province. Use Statistics Canada table 14-10-0201 — 17,598 series and ~5.3 million datapoints of monthly employment by industry, geography and employee type since January 2001.

1. Compare labor conditions across North America. Pair QCEW and OEWS with Statistics Canada SEPH tables; both sides arrive from Datadory as joined, attributed extracts.

The strongest support-services sources share four traits, and each decides whether a project survives contact with production:

What separates these records from a vendor-packaged product is convenience rather than coverage: the counts are the same near-census figures the agencies publish. The trade-off is that raw tables arrive unjoined — unless Datadory delivers them typed, keyed and joined.

Four kinds of teams drive most support-services data searches, and each starts from a different source:

Data scientists treat NAICS 56 as a feature store and modeling sandbox. Their queries look like "administrative support employment by county" or "canada payroll employment naics 56 monthly," and the answers live in QCEW's ~815,000-row county-year files and Statistics Canada's 17,598-series table 14-10-0201 — both delivered as join-ready tables. Our data scientists page collects these modeling workflows in one place.

Market researchers size service lines for go-to-market plans. They search phrases like "firm size distribution by industry data" and land on SUSB's firm, establishment, employment and receipts counts by six-digit NAICS and enterprise employment-size class, cross-checked against Quarterly Services Survey national revenue totals. The market researchers page maps the sizing workflow.

Vendor-risk and EHS teams screen facility-services contractors before awarding contracts, searching things like "osha injury data lookup by company." OSHA Enforcement & Injury Data serves them establishment-level filings current through August 2026 alongside inspection and citation history reaching back decades.

Sales and procurement teams prospect government business with queries like "federal contractor list by naics code," filtering the SAM.gov Entity Registry to janitorial, security and staffing NAICS/PSC codes — subsets measured in the tens of thousands of entities.

A practical support-services data stack stacks four layers, closest to the headcount question first:

1. UK extension layer. For cross-border work, join ONS BRES enterprise counts by SIC class to Companies House company-level status and incorporation dates — BRES is disclosure-controlled to base 5, so use Companies House for company-level precision.

Two hygiene checks belong in the design doc, not the retro: keep attribution strings for the sources that require them in your pipeline metadata, and never merge Statistics Canada information with other databases in a way that could identify an individual person, business or organization — its terms forbid it explicitly.

Every figure below comes straight from the cataloged source records, as of August 2026. Two carry the most planning weight: the QCEW archive's ~815,000 rows per recent-year file make county-level NAICS 56 modeling practical without a vendor middleman, and EDGAR's 2,486-filing hit count for a single 'staffing services' query shows how deep the corporate-filings corpus runs.

These are the questions buyers actually ask about diversified support services data, answered from the source records above.

Continue into the connected pages:

1. diversified-support-services data hub — the full 21-record pool on one index page, with quality scores, coverage notes, sample rows and field dictionaries for every dataset in this guide.
2. data scientists — how modeling teams rank all 1,744 cataloged datasets on schemas, granularity and training-permissive terms.
3. market researchers — sourcing guidance for client-ready market sizing built on firm-count and revenue series like SUSB and QSS.