Knowledge Graph optimization for GEO in 5 steps (2026)

What is a knowledge graph?

A knowledge graph is a database of the world's "things" and "concepts" (entities) and the relationships between them. Google's Knowledge Graph holds an enormous number of entities and serves as a foundation for how AI understands information and generates answers.

For example, the entity "Toyota" is linked to information such as the following:

• Type: automaker (Organization)
• Founded: 1937
• Headquarters: Toyota City, Aichi Prefecture
• CEO: Kenta Kon (since April 2026)
• Related entities: Lexus, Prius, Akio Toyoda, and others

If your brand is registered accurately in the Knowledge Graph, AI can recommend it based on correct information when it generates answers. Conversely, if the information in the Knowledge Graph is inaccurate or missing, AI answers may show incorrect information or may not mention your brand at all.

How Knowledge Graph optimization affects generative engine optimization (GEO)

Impact 1: AI recognizes your entity more accurately

When accurate information is registered in the Knowledge Graph, AI can more easily recognize your brand as a distinct entity. In particular, when other brands share your name or companies have similar names, Knowledge Graph information helps AI tell them apart correctly.

Impact 2: Your brand information stays accurate

When AI answers questions about your company and refers to Knowledge Graph information, accurate company details (such as founding year, location, and business description) are reflected in its answers.

Impact 3: Stronger connections to related entities

When the Knowledge Graph clearly shows how your brand relates to your products, services, and industry categories, AI becomes more likely to mention you for related queries.

Steps to optimize your Knowledge Graph presence

Step 1: Optimize your Google Business Profile

Google Business Profile (formerly Google My Business) is the most direct way to get your information into the Knowledge Graph.

1. Fill in every field: company name, address, phone number, business hours, categories, website URL, description, and more
2. Set categories accurately: choose the right primary and additional categories
3. Write a complete description: clearly describe your business, features, and strengths in 750 characters or fewer
4. Add photos regularly: keep your logo, storefront, and product photos up to date
5. Use posts: regularly post news, events, campaigns, and other updates

Step 2: Add your information to Wikidata

Wikidata is, in effect, the structured-data version of Wikipedia: an open knowledge base that many AI systems refer to.

What to register on Wikidata

• Company name (in both Japanese and English)
• Company type (e.g., kabushiki kaisha, the Japanese joint-stock company)
• Founding date
• Headquarters location
• Official website
• Industry and business description
• Social media accounts

Note: Wikidata requires "notability." Companies and brands that have been covered in news articles or other media are more likely to be accepted. If your notability is not yet sufficient, start by increasing your media exposure.

Step 3: Implement structured data

Implementing Schema.org structured data on your website lets you communicate entity information directly to search engines and AI.

At a minimum, implement the following structured data:

• Organization: basic company information
• WebSite: information about the site
• BreadcrumbList: the page's place in the site hierarchy

To strengthen the connection with the Knowledge Graph, it is important to list your external profile URLs comprehensively in the sameAs property.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Kabushiki Kaisha Sample",
  "alternateName": "Sample Inc.",
  "url": "https://example.co.jp",
  "logo": "https://example.co.jp/logo.png",
  "sameAs": [
    "https://twitter.com/sample_jp",
    "https://www.facebook.com/sample.jp",
    "https://www.linkedin.com/company/sample-inc",
    "https://www.youtube.com/@sample_jp",
    "https://www.wikidata.org/wiki/Q123456789"
  ]
}
</script>

For details on implementing structured data, see our guide to structured data and Schema.org for GEO.

Step 4: Keep your NAP information consistent

NAP (name, address, phone number) information is a basic element the Knowledge Graph uses to identify an entity. Make sure your NAP information matches exactly on all of the following platforms:

• Your own website
• Google Business Profile
• Social media profiles (X (Twitter), Facebook, LinkedIn, Instagram)
• Industry directory sites
• Job sites
• Press release distribution sites

Common mistake: variations in notation, such as "Kabushiki Kaisha X" versus the abbreviated "(Kabu) X" for a company name, or "1-1-1 Akasaka, Minato-ku, Tokyo" versus "Akasaka 1-chome 1-ban 1-go, Minato-ku" for an address. Each variant risks being recognized as a separate entity.

Step 5: Build relationships between entities

In the Knowledge Graph, relationships between entities also matter. Strengthen the connections between your entity and other entities in the following ways:

• Mention related entities on your own site: for example, state on a product page that it was "developed based on joint research with X University" and link to the university's website
• Get listed in industry association member lists: earn a link to your site from the association's website
• Exchange links with clients and partner companies: make the relationship explicit on both companies' websites

Knowledge Graph optimization checklist

1. Every field in your Google Business Profile is filled in
2. You have checked whether a Knowledge Panel appears for your brand
3. Your company is registered on Wikidata (or you have a plan to register it)
4. Organization structured data is implemented on your homepage
5. All your social media accounts are linked with the sameAs property
6. Your NAP information matches across all platforms
7. You have built connections between your entity and related entities

How to measure results

You can measure the results of Knowledge Graph optimization in the following ways:

• Search for your company name on Google and check the Knowledge Panel: check that the information is accurate and complete
• Ask AI about your company and check whether it answers accurately: ask "Tell me about X (your company name)" in Japanese and verify that the answer is accurate
• Track changes in your AI mention rate: use a tool such as AIOPulse to measure how your AI mention rate changes before and after your Knowledge Graph work

Summary

Knowledge Graph optimization is a foundational initiative for getting AI to recognize your brand accurately. Once it is set up properly, it keeps working over time, but update your information regularly.

To find out how AI currently recognizes your brand when asked in Japanese, and to decide how to prioritize your Knowledge Graph work, contact us.

Related articles

• GEO in Japan: measure the Japanese answers first (what GEO is and how to measure answers in Japanese)
• A complete guide to building authority for GEO (2026) (how E-E-A-T helps AI choose your brand)
• How brands get recommended in AI search (2026) (the 5 factors LLMs use to choose brands)
• Structured data for GEO: a Schema.org guide (2026) (practical steps for implementing Organization and sameAs)