Company Domain Finder: Name to Website, Logo, LinkedIn

Company Domain Finder: Company Name to Website, Logo & LinkedIn

Back on 2 November 2026 with new sources. It no longer reads any search engine. Names are matched against open data (Wikidata's official-website statements and, for Norwegian and Finnish companies, the national business registers) and against the domains the name spells, and every result is checked on the company's own home page. Your input and output fields are unchanged.

Turn a list of company names into official websites and domains. Paste names (with a country or city if you have one), and get back each company's domain, website URL, a confidence score, the method that found it, its logo URL and its LinkedIn company page. Every website is checked on the company's own home page before it is returned, so directories, news articles and parked domains do not slip through.

$0.80 per 1,000 domains found. Names that are not found are free.

What it does

For each company name, the Actor:

1. Looks the name up in open data. Wikidata (through the Wikimedia API, public-domain CC0 data) states the official website of many companies, with their country, logo and LinkedIn page. For names with country NO or FI, the Norwegian business register (Brreg, NLOD licence) and the Finnish Trade Register (PRH, CC BY 4.0) are asked for the website the company registered.
2. Tries the domains the name spells. Blue Bottle Coffee → bluebottlecoffee.com, Joe's Pizza, New York → joespizzanyc.com, plus your country's domain (.de, .co.uk, .com.tw and so on) and the get…/use…/…hq shapes. A guessed domain counts only when its own home page carries the company's name.
3. Verifies on the home page. It opens each candidate, follows redirects to the final domain (buttondown.email → buttondown.com), rejects parked and for-sale pages, and checks that the site's title, og:site_name or structured data names the company.
4. Scores and picks. Sources that agree raise the score; a page that does not carry the name lowers it. You get the best match, a 0 to 0.99 confidence, a level (high, medium, low) and up to 3 alternatives.

From the chosen home page it also reads the company's logo URL (structured data, header logo or touch icon), its LinkedIn company page and its X/Twitter, Facebook, Instagram and YouTube links.

Use cases

• Lead lists that only have names. Trade-show exhibitors, award lists, investor portfolios, CRM exports and spreadsheets of accounts: get the domain first, then the emails (see the pipeline below).
• CRM hygiene. Fill the missing website and domain fields on company records, with a confidence you can filter on.
• Enrichment input. Most enrichment tools key on a domain, not a name. This turns names into the key.
• Logos and LinkedIn pages for account lists, dashboards and pitch decks.
• AI agents and automations that need "the website of company X" as one clean API call.

Measured hit rate

Two fixed test sets, re-run on the Apify platform on 2 November 2026 on build 0.7.6 (run cf0APjxPKKDPbhi8c for the known companies, run fuTBHJZjM3nhh6Grq for the small and local businesses), default confidence floor 0.5. The first measurement on build 0.6.1 (runs 7nrHtP4mi552Ynb2n and LbcnwJsGRWIfks7AH) gave the same counts.

Known companies. 50 companies of mixed size from 20 countries (Germany, Japan, Taiwan, Brazil, India, Poland, Indonesia, Hong Kong and others):

Company size	Companies	Correct website	Wrong website	Not found (free)
Large (Siemens, Toyota, Petrobras...)	15	15 (100%)	0	0
Mid-size (Personio, KKday, Allegro...)	20	20 (100%)	0	0
Small (Buttondown, Monmouth Coffee, Tim Ho Wan...)	15	13 (87%)	0	2
All	50	48 (96%)	0	2

46 of the 48 were named by Wikidata and confirmed on the home page; 2 came from the name's own domain alone. Without the open-data step the same 50 names give 33 correct (66%). 47 of the 48 rows carried a logo URL and 34 a LinkedIn company page. The run took 87 seconds. The two misses are listed under a legal name that differs from their brand; search them by brand name instead.

Small and local businesses. 98 harder names: UK local firms from Google Maps lead lists, companies from job-board lists and leads that had no website. Of the 59 whose real site we know, 23 were found correctly (39%), 5 got a wrong site (82% of found rows right; two of the five were the company's site in another country, such as westelm.com for West Elm London), and 30 were not found (free). The 98-name run took 132 seconds and 31 websites were billed. Most local businesses have no open-data entry, so they rely on the name's own domain; expect roughly 3 to 4 in 10 found on such lists, and raise minConfidence to 0.8 when a wrong site costs more than a miss.

Wikidata is asked one request at a time, as the Wikimedia API rules require, so a long list of well-known names takes longer than before: plan on roughly 2 seconds per name at the default settings.

Correct means the returned domain is the company's own site (any of its official domains, such as bosch.de or bosch.com).

Input

{
  "companies": [
    "Stripe",
    "Spotify",
    "Plausible Analytics",
    "Gogoro, Taiwan",
    "Joe's Pizza, New York"
  ],
  "country": "",
  "industry": "",
  "minConfidence": 0.5
}

• companies: one per line. A location after a comma or pipe tells same-named companies apart: Bosch, Germany, Acme | Austin, US. Legal forms (Inc, GmbH, Ltd, S.A., K.K.) are fine.
• country / city: optional hints for every name without its own.
• industry: optional, e.g. software or coffee roaster, to prefer the right match among same-named companies.
• minConfidence: a row is returned as found (and billed) only at or above this. Default 0.5; use 0.8 for high precision.
• useWebSearch: retired on 2 November 2026 and ignored. The Actor reads no search engine; saved tasks that set it keep working.
• datasetId: read names from another Actor's dataset (fields companyName, company, organization, businessName, name or title). Optional per-row fields website, phone and postalCode (Google Maps Leads rows have all three) are used too: a website that loads and carries the company name is returned without a search (source: input-website), and a phone number or postal code printed on a candidate's home, contact or about page raises its confidence and replaces the city check.

Leave the input empty to run a built-in 3-company example.

Output

One row per company, in the order they finish:

{
  "input": "Gogoro, Taiwan",
  "companyName": "Gogoro",
  "found": true,
  "domain": "gogoro.com",
  "website": "https://www.gogoro.com/",
  "confidence": 0.99,
  "confidenceLevel": "high",
  "method": "wikidata+guess",
  "evidence": [
    "Wikidata Q19880591 (Gogoro: Taiwanese manufacturer of electric scooters and battery swapping infrastructure) lists it as the official website",
    "gogoro.com is registered and resolves",
    "home page names it ('Gogoro')"
  ],
  "homepageTitle": "Gogoro Global",
  "logoUrl": "https://cdn.gogoro.com/resources/pages/knowledge_graph_logo.png",
  "linkedinUrl": "https://www.linkedin.com/company/gogoro/",
  "socialLinks": {
    "twitter": "https://x.com/WeAreGogoro",
    "facebook": "https://www.facebook.com/gogoro",
    "instagram": "https://www.instagram.com/gogoro_taiwan/",
    "youtube": "https://www.youtube.com/channel/UChbmd1freKXd0VY0Qi-ibZg"
  },
  "wikidataId": "Q19880591",
  "description": "Taiwanese manufacturer of electric scooters and battery swapping infrastructure",
  "alternatives": [
    {
      "domain": "gogoro.io",
      "confidence": 0.25,
      "method": "guess"
    },
    {
      "domain": "getgogoro.com",
      "confidence": 0.18,
      "method": "guess"
    }
  ],
  "country": "TW",
  "city": null,
  "industry": null,
  "webSearchUsed": false,
  "webSearchBlocked": false,
  "scrapedAt": "2026-10-22T01:11:49Z"
}

A company that is not found (or is below your confidence floor) gets a free row with found: false, domain: null and its best guesses in alternatives, so you can see why.

Field	Meaning
domain	Registered domain, e.g. bosch.de (no www)
website	Home page URL after redirects
confidence	0 to 0.99. 0.8+ high, 0.6+ medium, below that low
method	Which sources named this domain, joined with +: wikidata, register (Brreg or PRH), guess (the name's own domain), input-website (instant-answer and search appear only in rows made before 2 November 2026)
evidence	Plain-English reasons, including the home-page check
logoUrl	Logo from the company's own site, else its open-data logo
linkedinUrl	LinkedIn company page linked from the site, else from open data
alternatives	Up to 3 other candidate domains with their scores
qualityFlags	Warnings on the returned site, empty when clean: placeholderPage, countryMismatch, cityNotOnSite, nameNotVerified, categoryName, inputWebsite, phoneMatch, postalMatch
localityEvidence	With a city input: where the site names that city (home page or a contact page), else null

Pricing

Pay per event: $0.0008 per company whose website is found ($0.80 per 1,000). Rows with found: false cost nothing, and there is no start fee. Apify platform usage is included.

Worked example: 2,000 names, 1,200 matched at the default confidence floor: 1,200 × $0.0008 = $0.96. The 800 misses are free.

Set Max total charge in the run options to cap any run.

From company names to emails: run it with Website Email Finder

Most buyers want the people's emails, not just the domain. Run the two Actors back to back, all on Apify:

1. Run Company Domain Finder on your names.
2. Run Data Gleaner's Website Email Finder & Contact Details Scraper (datagleaner/website-contact-details-scraper) with this run's dataset as input:

{ "datasetId": "<this run's default dataset ID>" }

It reads the website field of every found row and returns emails, phone numbers, social profiles and contact forms per site; rows with found: false have no website and are skipped. When a run finds at least one website, its status message and the NEXT_STEP record in its key-value store carry that input ready to paste.

To chain it automatically, open this Actor's Integrations tab, add "Run Actor" → datagleaner/website-contact-details-scraper, and set its input to {"datasetId": "{{resource.defaultDatasetId}}"}.

Cost for 1,000 names, 850 found, 600 with an email: 850 found-website charges (this Actor) plus 600 contact-scraper charges (its own price, below) come to about $1.88.

Need emails and phones too?

Set findContacts to true and the run does both steps in one go. After the websites are found, it runs Data Gleaner's Website Contact Details Scraper on your own Apify account for up to maxContactSites of them (default 50, at most 500, one per domain) and adds the result to each company's row: bestEmail, emailList, phoneList, contactForms and contactSocials. It is off by default; without it the run is exactly as before.

The contacts scraper bills its own price on your account, on top of this Actor's: its own per-site-with-email price (sites with no email are free). The run caps that sub-run at a small fixed amount per site sent. If the sub-run fails or runs out of time, you still get every domain result, and each row carries a contactsNote saying why it has no contacts.

{
  "companies": ["Stripe", "Spotify", "Plausible Analytics"],
  "findContacts": true,
  "maxContactSites": 50
}

Limits

• Companies with very common names and no location hint can match a same-named company; add a country or city, or raise minConfidence.
• For local businesses (a city is given), a site that never names the city on its home or contact page loses confidence, and directory, booking and marketplace sites are never returned as a company's site. Restaurants and shops with the same name in another city are the most common wrong match left; keep minConfidence at 0.6 or more for local lists.
• Some sites block automated visits; those are checked by name and domain only and score lower.
• Brand-new companies with no web presence yet are not found (and not billed).
• It returns one main site per company; subsidiaries and country sites can appear in alternatives.

FAQ

Does it need an API key or a login? No. It uses public open data, public web pages and the company's own home page.

What does "verified" mean? The candidate's own home page names the company (title, og:site_name or organisation structured data), it is not parked or for sale, and redirects are followed to the final domain.

Can I feed it a spreadsheet? Paste the name column into companies, or upload it as a dataset and pass datasetId.

How fast is it? 50 companies took 96 seconds on Apify. Open-data lookups go one at a time under the Wikimedia API rules, so 1,000 companies take roughly 30 to 40 minutes; set a run timeout to match.

Why was a well-known company not found? Usually a name shared by many companies with no hint. Add a country, city or industry and rerun just those rows; misses cost nothing.

Responsible use

This Actor reads open data about companies (Wikidata under CC0, Brreg under the Norwegian Licence for Open Government Data, PRH under CC BY 4.0, each named in a row's evidence) and the company's own home page, only where its robots.txt allows. It reads no search engine. It does not log in anywhere or collect personal data. You are responsible for how you use the results, including any obligations under data-protection and marketing laws where you operate.

Feedback

If this Actor saved you time, a rating on its Store page helps others find it. Problems or requests: open an issue on the Issues tab.