ExplainerAI visibility

Wikidata for Brands: A Step-by-Step Entry

by 3 min read

Short answerwikidata for brands

Wikidata is the structured database behind Wikipedia. Every item has an ID (Q-number), a label, a description, and a set of properties with values: instance of, founded, headquarters, official website, and so on. Answer engines and knowledge graphs use it to resolve entities, which is the step where a model decides that the "Acme" in a forum post and the "Acme" on your website are the same company.

A company does not need a Wikipedia article to have a Wikidata item. The notability bar is lower: the item needs to describe a real, identifiable entity and be useful for linking or referencing. Most established companies qualify. Here is how to create one properly.

Before you start

Search Wikidata for your company name first. An item may already exist, created by a bot from a company register. If it does, your job is to improve it, not duplicate it.

Gather your facts from the same source of truth your Organization schema uses: legal name, trading name, founding date, headquarters, founders, website, and any identifiers. Every value should be supportable with a public reference.

Step 1: Create the item

Log in with a Wikidata account. Use your own name and, on your user page, disclose that you are affiliated with the company. This is required under the paid-contribution policy and makes everything that follows easier.

Choose "Create a new item". Enter the label (the trading name, as it appears on your site), a short description ("American payroll software company" rather than "leading platform for teams"), and any aliases, including the legal name and common misspellings.

Step 2: Add the core properties

Add these in roughly this order:

  • instance of (P31): usually "business" or a more specific class like "software company".
  • official website (P856): the canonical homepage URL.
  • inception (P571): founding date, with precision set to year if that is all you can source.
  • headquarters location (P159): the city, as a linked item.
  • country (P17): the country of incorporation.
  • founded by (P112): one entry per founder, linked to their item if they have one.
  • industry (P452): the closest matching industry item.
  • legal form (P1454): corporation, LLC, GmbH, and so on.

Then the identifiers, which are the most valuable properties for disambiguation: Legal Entity Identifier (P1278), company register IDs, stock ticker if public, and social profile IDs (LinkedIn, X, GitHub, Crunchbase). Each one connects the item to a profile engines can cross-check.

Step 3: Add references

Every statement should have a reference: a URL to a page that supports it, with a retrieval date. Your own site is acceptable for basic facts. Independent sources are better where they exist. Unreferenced statements may be removed.

Add the new item's URL to the sameAs array in your Organization schema and to your llms.txt. Now your site points to Wikidata, Wikidata points to your site, and both point to the same LinkedIn and Crunchbase profiles. That loop is the point of the exercise.

Common mistakes

  • Marketing language in the description. It will be edited out, and it signals a promotional intent that invites scrutiny.
  • Creating items for products that are really just pages. A product item makes sense if the product has an independent identity and sources. A feature does not.
  • Entering facts that contradict your own site. A founding year that differs from your About page creates exactly the ambiguity the item was meant to remove.
  • Editing without disclosure. It is the fastest way to have your edits reverted and your account flagged.

Keeping it current

Review the item whenever a fact changes: a move, a new CEO, an acquisition. Put it on the same checklist as your schema and your LinkedIn page.

How we use it

At AEO GEO Labs, creating or correcting a Wikidata item is a standard step in the entity consistency pass, done after the Organization schema is finalized so the two match exactly. We also check that the item's description aligns with how we want engines to summarize the company, since in our testing that short description tends to show up in answers more often than its length suggests.

Do it this week

The whole process takes an afternoon for most companies. Start by searching Wikidata for your name, then run your site through the free schema markup generator at AEO GEO Labs to get a clean set of Organization facts to work from. Enter the same values in both places, reference each one, and link them together.

Keep reading

More on ai visibility

All posts

Rather not do it yourself? We'll get you named in AI answers.

Reading is the slow way. In 20 minutes we'll show you who ChatGPT recommends instead of you, and the three fixes we'd make first.

Book a call

Or get the free AI visibility report by email.

no strings, really