How we source, structure, enrich, and score 6,075,084 acquisition-ready businesses across 46 industries and 6 countries.
Buyers are actively searching this market. We are seeing active demand for companies like this. Check buyer interest privately for your business.
We are seeing active demand for companies like this.
Buyer names are not disclosed without permission; this proof uses anonymized demand category, size, geography, and criteria only.
English-speaking markets with the most lower-middle-market deal activity, each built from several independent datasets.
19 core verticals of 46 tracked: fragmented, recurring-revenue, owner-operated businesses common in PE and search fund deal flow. All are active in all 6 countries.
Records come from public registries, licenses, filings, open map data and business websites. Sources are free or budget-gated; treat fields as signals to review before outreach.
The Federal Motor Carrier Safety Administration publishes a census of every registered US motor carrier. Serava imports 1,048,269 active carrier records, 499,163 with a named owner or principal. Each record carries the operating address and phone from the federal filing.
The National Plan and Provider Enumeration System is the US registry behind every NPI number. Serava imports 968,197 provider records (dental, veterinary, optometry, physical therapy, chiropractic and more); 99.7% carry a named owner or authorized official, and 62.1% name a confirmed owner (the official whose surname is in the practice name). The name comes from the provider's own federal filing, so it is registry-grade, not scraped.
Secretary of State offices publish the registrations behind every legally formed entity. Serava imports 659,451 US entities, 228,326 with a named owner or officer. Where a state file names only the registered agent, Serava labels that person as the registered agent, never as the owner. They record entity type, registration date and the people behind the company, data that never appears on a website.
State licensing boards publish the licensees behind regulated trades and practices. Serava imports 330,972 licensed contractors, including 88,310 in Washington and Oregon where the file carries the owner's name, phone and licence tenure, plus 33,000+ named dental providers from state dental licence and provider listings. Licence tenure shows how long the licensee has held the licence or, for some dental listings, how long the provider has been listed. Tenure is shown only where Serava has a reliable start date.
A baseline open-map layer for visible local operators. Serava queries Overpass where OSM tagging is dense, then fills thin markets with official registries, licenses, filings and demand-led imports. OSM is not treated as complete for niche categories.
Companies House publishes a monthly bulk CSV of all registered UK companies. Serava keeps active companies incorporated 3+ years ago with a valid postcode, maps SIC 2007 codes to our industry types, and geocodes them via postcodes.io. Each import cycle adds 100,000 to 300,000 UK businesses with registration data OSM lacks.
The New Zealand Companies Office publishes the official Companies Register as monthly bulk data. Serava imports 395,696 NZ companies, 388,299 (98.1%) with a named director from the register's director records (the earliest-appointed director on file), plus 62,607 registered websites. It covers every region from Northland to Southland. The company list is the April 2026 import; director names and websites are updated monthly from the official feed.
The Open Database of Businesses (ODBus) is Statistics Canada's public extract of the National Business Register: names, addresses, NAICS codes and employee size class for most Canadian employer businesses. Serava maps its NAICS codes to our taxonomy, filling gaps where OSM tagging is thinner, such as Quebec and the Maritimes.
The City of Montréal publishes an open-data file of every commercially registered business in the city, including small operators missing from national datasets. Serava imports it from the open data portal and geocodes each record.
The US Small Business Administration releases historical 7(a) loan data under FOIA: borrower name, city, state, NAICS code, loan amount and jobs retained. Serava cross-references it to append employee estimates and revenue proxies. SBA loans skew toward established, owner-operated businesses, so this is a strong acquisition-relevance signal.
SAM.gov is the US System for Award Management, listing businesses registered for federal contracts. Registrations can include NAICS codes, employee counts, annual revenue and owner demographics. Serava uses matching records to add self-reported employee and revenue signals.
TDLR publishes daily CSV license files for Texas contractors, facilities, towing and vehicle storage, professional employer organizations, EV supply providers, mold companies and related businesses. Serava imports active business-level records, maps them to our taxonomy, and places them by ZIP or county, depending on each file's detail.
Registry data says a business may exist and where. Enrichment adds contact, size, source and reputation signals when available. Paid sources stay budget-gated and demand-led; public imports and website crawls run first.
Budget-gated machine extraction from public web signals, reserved for demand-led queues, paid target maps or reviewed expansion, not broad enrichment.
Optional paid contact and reputation signals from Google's business index, matched by name and coordinates when budget allows. It can be expensive, so Serava runs it only on high-intent target maps, not the whole dataset.
Free, unlimited SPARQL queries against Wikidata's knowledge graph, useful for notable companies whose founder, CEO, headcount and revenue are recorded. Serava queries it for any company matching a known entity.
IRS Form 990 data for US nonprofits, associations and foundations: total revenue and assets from the latest filing. Useful for associations, healthcare nonprofits and trade schools in the home services and healthcare categories.
Each company is graded by how much real data backs it. Modeled estimates never inflate the grade.
Each company gets a composite fit score from 0 to 100, built to surface traits buyers may care about. Higher scores rank first by default. The fit score is a model estimate for sorting, not a measurement. For most companies its size input is a placeholder, and for many its age is too.
Known years in business read as a succession signal: 15+ years ranks highest. Serava has no owner tenure data, and a company with no known founding date earns no points here.
Total operating age. Older businesses have survived multiple economic cycles and carry lower go-forward risk.
Where Diffbot, SBA or SAM data exists, revenue is scored against the mandate's range. The centre of the range scores highest.
Businesses with phone, website, contact-page or Google Places signals score higher, which can correlate with active operations and reachability.
Google rating and review count signal satisfaction and market presence. High rating at high volume points to a stable customer base.
Sources overlap, so deduplication runs at the database layer. Each company is keyed on its normalized name plus rounded latitude and longitude. When a business appears in both OSM and a national registry, the records merge rather than duplicate, and the second source's extra fields are appended.
Companies are re-graded as verified data lands. Registry re-imports follow each source's release schedule, and enrichment runs on higher-intent companies first.
Map the market against your acquisition criteria.
Build your mapDescribe your thesis and see how many owner-led businesses match across 5.8M companies, free.
Find your targets free