# Get a company name from a domain for free: markup, Wikidata, EDGAR

Enrichment Desk. A ZeroClick publication (https://enrichmentdesk.com/about/). Published 2026-11-02.
Format: Guide. Topic: Company data (https://enrichmentdesk.com/topics/company-data/).
Canonical: https://enrichmentdesk.com/articles/domain-to-company-name/

Ask Wikidata which item has https://instagram.com as its official website, and it answers JKL Dolphins Kampala; ask for https://www.instagram.com/ and it answers Instagram. A domain points to a website, not to a company, so this guide takes the company from sources that state it on purpose: the site's own markup and terms, Wikidata and the SEC's EDGAR. Every domain ends with a status your agent can write back, and every disagreement is flagged. You need an HTTP client, a list of domains or work-email addresses, and a User-Agent header that names your agent and gives a contact address.

## How to get a company name from a domain

1. Take the domain from each address (the part after the last @, in lowercase) and check it against the open-source [free-email-domains](https://github.com/Kikobeats/free-email-domains) list, which is MIT-licensed. A match means the status is `free mail` and the work stops.
2. Fetch the homepage and read its canonical link and any [schema.org Organization](https://schema.org/Organization) block, keeping `name`, `legalName` and `url`.
3. Where the markup gives no `legalName`, open the terms page or footer and take the company it names as the operator.
4. Query Wikidata's [official website property (P856)](https://www.wikidata.org/wiki/Property:P856) with all eight forms of the domain: http or https, with or without www, with or without a trailing slash. Send a User-Agent that names your agent and gives a contact address, as Wikimedia's [User-Agent policy](https://foundation.wikimedia.org/wiki/Policy:Wikimedia_Foundation_User-Agent_Policy) asks; the [query service manual](https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual) says clients that ignore the policy may be blocked completely.
5. Accept a Wikidata item only when a single item answers across all eight forms and it agrees with the site's own name.
6. For a US-listed company, search EDGAR for the legal name, then read its [submissions record](https://www.sec.gov/search-filings/edgar-application-programming-interfaces) by CIK, the SEC's [permanent filer number](https://www.sec.gov/os/accessing-edgar-data), for the registered name, ticker and exchange.
7. Look the domain up in [RDAP](https://about.rdap.org/) to confirm it is registered, and expect to find the registrar there, not the owner.
8. Write back one status with every name found and its source: `confirmed`, `trade name` (the brand and the legal entity behind it, with both names), `free mail`, `refused` (a source declined the request) or `not found`. When sources name different entities, keep them all and flag the domain.

## Example calls

Wikimedia asks scripts for an informative User-Agent with contact information, and the SEC asks EDGAR users to declare one too. The requests below assume one header in Wikimedia's format:

```
User-Agent: domain-resolver/1.0 (https://example.org/resolver; ops@example.org)
```

Our fetch of Stripe's homepage returned a canonical link to https://stripe.com/, the form of the domain the site uses for itself, and an Organization block in its markup. Trimmed to the identity fields:

```
GET https://stripe.com/

<link rel="canonical" href="https://stripe.com/"/>
<script type="application/ld+json">
{"@type": "Organization", "url": "https://stripe.com/",
 "name": "Stripe", "legalName": "Stripe, LLC"}
</script>
```

`name` is the brand and `legalName` the registered company, the split that step 8's statuses turn on. The Organization block is where a site names itself on purpose. Schema.org defines `legalName` as the official name of the organization, such as the registered company name, and `name` as the name of the item. [Google](https://developers.google.com/search/docs/appearance/structured-data/organization) recommends placing the block on the homepage or an About page, with `legalName` where it differs from `name`. Google lists no required properties, and schema.org's usage figures put Organization on more than 10 million domains but [legalName](https://schema.org/legalName) on 1 to 10 million, so step 3 often does the work. Instacart's terms page is an example: it names Maplebear Inc., a Delaware corporation doing business as Instacart, as the company behind the service.

Wikidata comes next, for known brands. P856 holds a full URL, current or former, and a query matches only that exact string. Our query sent eight forms of stripe.com to [Wikidata's query service](https://www.wikidata.org/wiki/Wikidata:SPARQL_query_service) in one request:

```
GET https://query.wikidata.org/sparql?query=<the query below, URL-encoded>&format=json

SELECT ?url ?item ?itemLabel WHERE {
  VALUES ?url { <https://stripe.com> <https://stripe.com/>
    <https://www.stripe.com> <https://www.stripe.com/>
    <http://stripe.com> <http://stripe.com/>
    <http://www.stripe.com> <http://www.stripe.com/> }
  OPTIONAL { ?item wdt:P856 ?url . }
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en,mul". }
}
```

One form found the company, and seven found nothing. The response, trimmed:

```
{"url": {"type": "uri", "value": "https://stripe.com/"},
 "item": {"type": "uri", "value": "http://www.wikidata.org/entity/Q7624104"},
 "itemLabel": {"xml:lang": "en", "type": "literal", "value": "Stripe"}}
{"url": {"type": "uri", "value": "https://www.stripe.com"}}
... six more bindings with no item
```

Querying P856 with four forms of instagram.com named two different entities.

```
{"url": {"value": "https://instagram.com/"}}
{"url": {"value": "https://www.instagram.com/"},
 "item": {"value": "http://www.wikidata.org/entity/Q209330"},
 "itemLabel": {"xml:lang": "mul", "value": "Instagram"}}
{"url": {"value": "https://instagram.com"},
 "item": {"value": "http://www.wikidata.org/entity/Q140386630"},
 "itemLabel": {"xml:lang": "en", "value": "JKL Dolphins Kampala"}}
{"url": {"value": "https://www.instagram.com"}}
```

Two forms of one domain naming two entities is a flag, not an answer. Instagram's label came back under the code mul, not en. The manual says that when an item has no label in any requested language, the query returns the bare Q-id instead, which is why the query asks for en,mul. P856's own page marks its values as ones that sometimes change, so the write-back carries the date checked.

EDGAR then confirms the legal entity behind a US-listed company. A company search for Maplebear, the name Instacart's terms gave, returned one filer.

```
GET https://efts.sec.gov/LATEST/search-index?keysTyped=Maplebear

{"hits": {"total": {"value": 1, "relation": "eq"},
  "hits": [{"_id": "1579091",
    "_source": {"entity": "Maplebear Inc. (CART)", "tickers": "CART"}}]}}
```

The submissions record for that CIK gave the registered name, ticker, exchange and state of incorporation. Its website field was empty, and it listed no former names.

```
GET https://data.sec.gov/submissions/CIK0001579091.json

{"cik": "0001579091", "entityType": "operating", "name": "Maplebear Inc.",
 "tickers": ["CART"], "exchanges": ["Nasdaq"], "stateOfIncorporation": "DE",
 "website": "", "formerNames": []}
```

EDGAR holds the entity, not the domain. The link from a domain to its filer has to come from the site itself.

RDAP, the Registration Data Access Protocol that replaces WHOIS, named only the registrar for stripe.com.

```
GET https://rdap.org/domain/stripe.com

{"ldhName": "STRIPE.COM",
 "entities": [{"roles": ["registrar"],
   "publicIds": [{"type": "IANA Registrar ID", "identifier": "447"}],
   "vcardArray": ["vcard", [["version", {}, "text", "4.0"],
     ["fn", {}, "text", "SafeNames Ltd."]]]}],
 "events": [{"eventAction": "registration", "eventDate": "1995-09-12T04:00:00Z"},
   {"eventAction": "expiration", "eventDate": "2027-09-11T04:00:00Z"}]}
```

The instacart.com record did the same: it named Amazon Registrar, Inc. and no owner.

That gap is built into the rules. ICANN's [Registration Data Policy](https://www.icann.org/resources/pages/registration-data-policy-2024-02-21-en), in effect since 21 August 2025, lists the registrant's name among the fields to publish but subjects it to the policy's redaction rules. Accredited registrars and registries must redact the name where law requires it, and may redact it whenever they have a commercially reasonable purpose. The registrant organization is published only when the holder confirms the value and agrees to its publication. RDAP confirms that a domain is registered, but it cannot be counted on to say who runs it.

Put together, stripe.com is `confirmed`: its markup and Wikidata both name Stripe, and the markup adds the legal name Stripe, LLC. instacart.com is written back as `trade name`, with both Instacart and Maplebear Inc., which EDGAR confirms as a Nasdaq filer. instagram.com is flagged, with both Wikidata items kept.

## What it costs

All eight calls cost $0. A domain takes three to six calls:

- the homepage, Wikidata and RDAP every time;
- the terms page when the markup has no legal name;
- two EDGAR calls for a US filer.

The free-mail check is a lookup in a downloaded file, not a call.

| Call | Price per call | Limit to plan around |
|---|---|---|
| Homepage, terms page | $0 | Set by each site |
| Wikidata query service | $0 | 60 seconds of query time per minute per client, 30 error queries a minute, 5 parallel queries per IP, then 429 with Retry-After; User-Agent policy: clients without a descriptive one may be blocked |
| EDGAR search and submissions | $0 | 10 requests a second; declared User-Agent; no API key |
| RDAP via rdap.org | $0 | 10 requests in 10 seconds, then 429 |
| free-email-domains | $0 | MIT licence; validated against DNS every day |

rdap.org's limit caps a run at about one domain a second. For heavier use, rdap.org points to an RDAP client that reads the registry bootstrap files itself, and EDGAR publishes a nightly bulk file of every submissions record. Past that point, paid domain-to-company services priced per call take over. Zero, the search engine for agent services run by ZeroClick, which publishes this site, lists them: [search it for domain-to-company services](https://www.zero.xyz/search?q=domain+to+company).

## When this is the wrong tool

- **The site says nothing about itself.** A small private firm with no Organization markup and no operator in its terms becomes `not found`, and Wikidata and EDGAR may not list it either.
- **The company doesn't file with the SEC.** EDGAR confirms legal entities only for its filers. Outside them, the legal name rests on the site's own statement.
- **Proof of control.** None of these sources proves who controls a domain, and the RDAP records here named only registrars.
- **Names, not domains.** If the agent holds company names and needs domains, our [companion guide](https://enrichmentdesk.com/articles/find-company-domain-from-name/) runs the reverse.
- **Large runs in 2026 and 2027.** The Wikidata Query Service is moving to a new backend over those two years, with new endpoint URLs that may require existing queries to be revised.

## Frequently asked questions

### How do I get a company name from a domain for free?

Read the homepage's schema.org Organization markup and the terms page, then match all eight URL forms of the domain against Wikidata's official website property. For a US-listed company, EDGAR's APIs confirm the registered name and need no key.

### How do I get a company name from an email domain?

Take the part after the last @ and check it against a personal-mail list such as free-email-domains, which holds over 30,000 free and disposable domains and matches them exactly. Resolve only the domains that are not on the list.

### Can I get a company's legal name from its domain?

Often from the site itself: schema.org's legalName field and the operator named in the terms both state it. For a US-listed company, EDGAR's submissions record gives the registered name, though Maplebear's record left its website field empty.

### Does WHOIS or RDAP show who owns a domain?

Not reliably. ICANN's Registration Data Policy lets registrars redact the registrant's name, and the RDAP records for stripe.com and instacart.com named only their registrars.

### Why does Wikidata return the wrong company for a domain?

Its official website property matches exact URL strings. That is how https://instagram.com found JKL Dolphins Kampala while https://www.instagram.com/ found Instagram.
