What is Wikidata?

Wikidata is a free, multilingual knowledge base run by the Wikimedia Foundation that stores facts about people, companies, places, and concepts as structured, machine-readable statements. Every entry carries a permanent identifier with the prefix Q, such as Q2013 for Wikidata itself. Search engines and AI systems use these identifiers to recognize an entity unambiguously and to cross-check facts about it against other sources.

Term profile at a glance

How is a Wikidata entry structured?

A Wikidata entry consists of an identifier, a description in several languages, and a list of statements, each backed by a source. Wikidata calls such an entry an item. It is made up of individual facts that a program can read without interpretation.

An item has 5 building blocks:

• QID: The identifier made of the letter Q and a number. It stays the same even when the name or description changes, which makes it the fixed anchor for the entity.
• Label, description, and aliases: The name, a short classification, and alternative spellings, kept separately for each language.
• Statements: Each statement connects a property with a value. Properties carry the prefix P, such as P856 for the official website.
• Qualifiers and references: Qualifiers refine a statement (for example with a time period), references back it up with a source.
• Sitelinks: Links to the matching articles on Wikipedia and other Wikimedia projects, where such articles exist.

For a company, these properties are the most common:

The value of a statement is often an item in its own right. The statement “headquarters location: Munich” points to the item for the city of Munich with its own QID. These connections form a knowledge graph in which every entity is described through its relationships to other entities. In early 2025, Wikidata held more than 1.6 billion such statements.

Why does a company need a Wikidata entry?

A Wikidata entry makes a company unambiguously identifiable for machines and provides an independent, openly readable confirmation of its core facts. That pays off in 4 places:

• Disambiguation: When several companies or people share a similar name, the QID reliably separates them. Search systems no longer have to guess from context which company is meant.
• Knowledge Graph and knowledge panel: Wikidata is one of the structured sources from which Google takes facts for its Knowledge Graph. In December 2014, Google announced that it would shut down its own database Freebase and move its content to Wikidata. An entry strengthens the data available for a knowledge panel, but it does not guarantee one.
• AI answers: Wikidata is freely licensed and fully accessible by machine. Its data therefore flows into many datasets that language models work with, and into systems that look up facts at answer time. An LLM that knows a company by its QID is less likely to confuse it with a namesake.
• Consistency: Name, headquarters, founding year, and website sit in one place that other sources reference. When the website, directories, and Wikidata contradict each other, trust in all three drops. The same logic applies to NAP data in local search.

A Wikidata entry is therefore one building block of entity work. It takes effect together with structured data on your own website, consistent profiles, and brand mentions in independent sources. This interplay is the core of GEO, optimization for AI visibility.

Which companies are allowed to create a Wikidata entry?

Any company that can be described unambiguously with serious, publicly available sources can have a Wikidata entry. Wikidata’s notability policy names 3 criteria, and meeting one is enough:

1. Sitelink: The item has an article on Wikipedia or another Wikimedia project.
2. Verifiable entity: The item is a clearly identifiable entity that can be described using serious and publicly available references.
3. Structural need: The item is needed to make statements in other entries useful, for example as the employer of a person who is already recorded.

For most mid-sized companies, the path runs through the second criterion. Suitable references include the commercial register entry, authority identifiers such as an ISNI, press coverage, industry directories, and chamber or association memberships. Your own website alone rarely counts as a reference, because it is not an independent source.

The community decides on disputed entries through deletion requests. Entries that read like advertising or whose statements lack a reference are regularly deleted. A Wikidata entry therefore stays sober, factual, and referenced in every statement.

How do you create a Wikidata entry for a company?

A Wikidata entry is created in 6 steps, from the duplicate check to the link back from your own website. The biggest effort is gathering the references.

1. Check for duplicates: Search wikidata.org for the company name, former names, and spelling variants. If an entry already exists, you extend that one.
2. Create a user account: A registered account makes edits traceable and comes across as more trustworthy to the community than an anonymous IP address.
3. Create a new item: Use “Create a new Item” to enter the label, description, and aliases, at least in German and English. The description is short and neutral, for example “search engine optimization company from Munich”.
4. Add statements with references: Type of organization, official name, inception, headquarters, website, and industry. Every statement gets a reference, for example the URL of the register entry.
5. Link identifiers: Add external identifiers such as ISNI, OpenCorporates, or the LinkedIn company page as separate statements. Each identifier connects the entry with another source that describes the same entity.
6. Link back from your website: Add the QID to the Organization schema on your own website as sameAs. Only this link back closes the loop between your website and Wikidata.

This is what the link back looks like in JSON-LD, with placeholders for the domain and the QID:

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://www.example.com/#organization",
  "name": "Example Inc.",
  "url": "https://www.example.com/",
  "sameAs": [
    "https://www.wikidata.org/wiki/Q999999999",
    "https://www.linkedin.com/company/example/"
  ]
}

After that, the same rule applies as for any source about your company: When the headquarters, name, or website changes, update the entry. An outdated Wikidata entry contradicts your own website and weakens exactly the signal it is meant to strengthen.

What is the difference between Wikidata and Wikipedia?

Wikipedia explains a topic in prose, Wikidata stores the facts behind it as individual, language-independent records. Both projects belong to the Wikimedia family and are closely connected: Many Wikipedia infoboxes pull their values directly from Wikidata.

For a company without a Wikipedia article, Wikidata is the realistic way into the Wikimedia world. A Wikipedia article cannot simply be created, because the German-language Wikipedia only includes companies with high figures such as headcount and revenue or with outstanding significance. A Wikidata entry only requires that the facts can be verified.

Frequently asked questions about Wikidata

Does a Wikidata entry cost anything?
No. Creating and editing entries is free, and anyone with internet access can contribute. Costs only arise when a service provider takes over researching the references and maintaining the entry.

Do I need a Wikipedia article to get a Wikidata entry?
No. A sitelink to Wikipedia is only one of 3 ways to meet the notability criteria. A company that can be described unambiguously with serious, public sources meets them without an article as well.

Can I edit my company’s Wikidata entry myself?
Yes. Wikidata is freely editable, including by the company itself. Every change appears in the revision history, and the community reviews it. Stick to neutral, referenced facts, and the entry will stay.