Franchise & Multi-Location Finder — Chains by Category

Type a business category and a region — pizza in Texas, gym in Ontario, dental clinic in UK — and get the chains and franchise brands operating there, one row per brand: a verified location count (with its source), the countries and states covered, HQ address and phone, whether it is a franchise (with the fee, minimum cash and total investment range when a franchise directory states them), founding year, emails and social profiles, sample locations and a who-to-sell-to hint. Or paste a list of brand names to verify.

No login, no API key, no third-party Actors. Dataset-only output, MCP-ready. Pay per brand delivered.

Who it is for

• Enterprise and mid-market local sales — POS, loyalty, ordering, scheduling, payroll, insurance, signage, cleaning, packaging: anyone who sells to the system, not the store. One row = one HQ to call, with the size of the footprint it controls.
• Franchise suppliers and consultants — brands that franchise, with fee / investment economics and where they are expanding.
• Multi-location marketing and local SEO agencies — chains by category and region, ranked by size.
• Market researchers — how many pizza chains operate in Texas and how big each one is.

What it does

1. Discovers brands on Google Maps: <category> in <region> (several result pages), then the busiest cities in the results get their own search so regional chains surface too. Places are folded into brands by website domain (or normalised name when a place has no website) and ranked by how often they showed up.
2. Verifies the location count on the brand's own site: sitemap URLs under /locations/, /stores/, … (or the locations. subdomain many chains use), JSON-LD / embedded store lists on the locator page, and the brand's own statement ("more than 5,800 clubs"). Wikipedia's infobox ("Number of locations") and the franchise directories' unit counts are read too; the row keeps all of them and names the one it chose.
3. Enriches each brand: HQ address and phone (JSON-LD Organization, contact / about pages, footer), LinkedIn, franchise page, founded year, employees; Franchise Direct, Franchise Gator and the IFA directory for franchise fee, minimum cash, total investment range and unit count; emails and social links from the website.
4. Delivers brands with at least minLocations locations (and franchise evidence when franchiseOnly), charging per brand — never for candidates below the minimum, duplicates or brands without a verifiable count.

Why this Actor

• Brand rows, not place rows. Google Maps scrapers give you 300 pizza places; this gives you the 15 pizza chains behind them, each with one HQ and one footprint number.
• Counts you can defend. locationsCountSource says whether the number came from the brand's store locator, Wikipedia, a franchise directory, the brand's own statement or only the Maps sample — and locationCounts shows all of them side by side.
• Franchise economics where they exist. Fee, minimum cash, investment range and franchising-since from the directories that publish them; null where they do not.
• Nothing guessed. Every field is read from a page the row links to (sources[], locatorUrl, wikipediaUrl, directoryUrls).

Output fields

Field	Description
brand, domain, website	Brand name (most frequent normalised name among its places), registrable domain (the row key), website
category, categories, mapsCategories	Your category that surfaced the brand; all of yours that did; Google Maps categories seen
locationsCount, locationsCountSource, locationsCountDetail	The chosen count, its source (locator / wikipedia / directory / site-claim / maps-sample) and how it was read
locationCounts	{ mapsSample, locator, wikipedia, directory, siteClaim } — every count that was found
countries, states, regionsSearched	Countries (ISO) and states / provinces seen in the Maps sample and HQ; the regions you searched that surfaced the brand
hqAddress, hqCity, hqState, hqCountry, hqPhone	Headquarters from the brand site (JSON-LD / contact page / footer), Wikipedia or a directory; HQ phone from the site
isFranchise, franchiseEvidence, franchisePageUrl	true with evidence (directory listing, franchising page, Wikipedia), else null
franchiseFeeUsd, minCashUsd, investmentMinUsd, investmentMaxUsd, franchisingSince	From Franchise Direct / Franchise Gator / IFA, as stated there
foundedYear, employees, description	Wikipedia infobox / JSON-LD / directory; site meta description
emails, primaryEmail, phonesFromWebsite, socials, linkedinUrl	Website contacts (charged only when found)
locatorUrl, wikipediaUrl, directoryUrls	The pages the facts came from
avgRating, totalReviewsInSample, sampleLocations[]	From the Maps sample: average rating, summed reviews, up to 5 places (name, address, phone, rating, Maps link)
sellTo	Who to sell to, e.g. "5,463-location chain headquartered at …; sell to the franchisor HQ (VP Franchise Development / Operations, CMO, IT / POS lead) …"
sources[], scrapedAt	Which sources contributed; ISO timestamp

Also in the key-value store: BRANDS.csv (spreadsheet-ready, first 5,000 rows) and OUTPUT (run summary: brands per category, count sources, candidates found / enriched / dropped with reasons, Maps pages, transport stats).

How to use

1. Enter one or more categories (pizza, gym, dental clinic, car wash) and regions (Texas, Ontario, UK). Or paste brands to verify (Jersey Mike's, anytimefitness.com).
2. Set minLocations (default 5) and maxBrands (default 50). Turn on franchiseOnly for franchisors only.
3. Run. Rows land in the dataset as each brand is verified; export as CSV / JSON / Excel or read them through the API and MCP.

Tips: a region-level search returns 60–80 places; the city follow-ups add ~250 more, so 30–40 brands are usually seen twice or more per category × region. Big national chains appear in every category run; regional chains (10–100 locations) are the ones you will not find elsewhere. For an exact worldwide figure look at locationCounts.wikipedia; for the US locator count look at locationCounts.locator.

Time budget: the run stops starting new work at the earlier of its own 50-minute ceiling and the run timeout minus 40 s, delivers the brands verified by then and names the cut in the status message and in OUTPUT.timeBudget (cutShort, candidatesUnverified). Once maxBrands rows are delivered the run ends without waiting for brands still being enriched (they are not charged). With less than 10 minutes on the clock, discovery reads 2 Maps pages per query and 3 follow-up cities instead of 4 and 5. The form prefill (pizza in Texas, 6 brands) finishes in about a minute at 1024 MB.

Cost

Pay per event: $0.05 per brand delivered, $0.01 per brand with contacts (an email or social profile was found), $0.005 per Actor start. Platform usage is on top and small: a 15-brand run costs about $0.01–0.02 of compute.

Input example

{
    "categories": ["pizza", "gym"],
    "regions": ["Texas", "Ontario"],
    "minLocations": 5,
    "maxBrands": 40,
    "franchiseOnly": false,
    "enrichContacts": true,
    "brands": ["Jersey Mike's", "anytimefitness.com"]
}

Output sample

{
    "brand": "Cicis Pizza",
    "domain": "cicis.com",
    "website": "https://www.cicis.com/",
    "category": "pizza",
    "locationsCount": 499,
    "locationsCountSource": "locator",
    "locationsCountDetail": "499 sitemap URLs at depth 3 under \"locations\" (612 matched in total)",
    "locationCounts": { "mapsSample": 6, "locator": 499, "wikipedia": 279, "directory": null, "siteClaim": null },
    "countries": ["US"],
    "states": ["Texas"],
    "hqAddress": "1080 W Bethel Rd, Coppell, TX 75019",
    "hqPhone": "(972) 745-4200",
    "isFranchise": true,
    "franchiseEvidence": "listed on franchisegator; franchise page on the brand site (https://www.cicis.com/franchising)",
    "franchiseFeeUsd": 30000,
    "investmentMinUsd": 610000,
    "investmentMaxUsd": 1050000,
    "foundedYear": 1985,
    "emails": ["franchising@cicis.com"],
    "sellTo": "499-location regional chain headquartered at 1080 W Bethel Rd, Coppell, TX 75019; sell to the franchisor HQ (VP Franchise Development / Operations, CMO, IT / POS lead) — franchisees buy locally, the system buys centrally.",
    "sources": ["google-maps", "brand-site", "brand-site-locator", "wikipedia", "franchisegator", "brand-site-contacts"],
    "scrapedAt": "2026-09-26T15:03:20.000Z"
}

FAQ

Why do the counts differ between sources? A brand's US site lists its US locations; Wikipedia gives the worldwide figure of a given year; a directory reports the units it was told. The row keeps all of them in locationCounts and locationsCountSource names the one in locationsCount (store locator first, then Wikipedia, directory, the brand's own statement, and the Maps sample as a lower bound — a clearly larger source wins so a partial sitemap never undercuts the others).

A brand I know is missing. Google Maps ranks independents high on region-level queries; raise mapsPagesPerQuery / citiesPerRegion, add the brand to brands[], or search a narrower region. Brands without any verifiable count are listed in OUTPUT.dropped with the reason.

Is isFranchise: null a "no"? No — it means no evidence was found. Only true is asserted.

Which regions work? Anything Google Maps understands: US states, Canadian provinces, UK counties / metros, countries, cities. Google's country setting follows the region (Texas → US, Ontario → CA, UK → GB).

Does it need proxies? Every site is read through the run's own connection first (Google Maps included); the proxy you configure is the per-host fallback. If Maps starts answering with captchas, use RESIDENTIAL.

Legal. Reads public pages only: Google Maps search results, brand websites, Wikipedia and public franchise directory listings. Use the data in line with the sources' terms and your local law.