Knowledge Graph SEO

TL;DR — The Knowledge GraphThe Knowledge Graph is Google's database of entities — people, places, organizations, and things — and the factual relationships between them. It's separate from any single website's structured data: your schema markup is one of many possible inputs to the graph, not the graph itself. is Google’s giant database of things — people, companies, places — and the facts about them. It’s not the same as the schema markup on your website; your markup is just one thing that might feed it. The box that pops up on the right of search results with a photo and quick facts is a Knowledge Panel — the visible part of that database. You can sometimes claim and correct a panel, but you can’t force one to exist, and nothing guarantees you get one.

What the Knowledge Graph is

Google describes the Knowledge Graph as a system for understanding facts about people, places, and things and their relationships. Evidence for this claim Official or primary documentation supporting the adjacent article claim, with scope limited to the source's published description. Scope: No ranking guarantee or undisclosed system mechanics are inferred beyond the cited source. Confidence: high · Verified: 2026-07-14 Google: Introducing the Knowledge Graph A Knowledge Panel is a search-result presentation generated from multiple sources, not a feature guaranteed by adding schema. Evidence for this claim Official or primary documentation supporting the adjacent article claim, with scope limited to the source's published description. Scope: No ranking guarantee or undisclosed system mechanics are inferred beyond the cited source. Confidence: high · Verified: 2026-07-14 Google: Knowledge panels

When you search for a well-known person, company, or place, Google often shows a box with a photo, a short description, and quick facts — birthday, founder, headquarters, and so on. That information doesn’t come from any one website. It comes from Google’s own database of entities and the facts connecting them. That database is the Knowledge Graph.

Google launched it back in 2012 with a memorable line: it was built to understand “things, not strings.” Before that, search mostly matched the words you typed against words on pages. The Knowledge Graph lets Google understand that “Apple” the company is a different thing from “apple” the fruit, and that a company has a CEO, a founding date, and products — real facts about a real thing.

The one thing people get wrong

Here’s the confusion I see most: people think the Knowledge Graph is their website’s structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. (schema markupSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor.). It isn’t.

• Your structured data lives on your pages. It’s a way of labelling your content so machines can read it more easily.
• The Knowledge Graph lives at Google’s level. It’s Google’s database, built from hundreds of sources.

Your structured data is one possible input to the graph — a signal Google might use to understand and disambiguate your entity. But it’s not the graph, and adding markup doesn’t drop your facts straight into Google’s database.

One more distinction worth naming: some SEO platforms and agencies talk about “building a knowledge graph” for your own site — an internal map of your content’s entities and relationships. That’s a legitimate practice for organizing content and helping systems parse it more easily. But it’s still your graph, sitting on your own infrastructure. It isn’t Google’s Knowledge Graph, doesn’t automatically feed into it, and building one doesn’t guarantee Google ingests or surfaces any of it.

What a Knowledge Panel is (and how to get one)

The Knowledge Panel is the visible box. It’s the part of the Knowledge Graph you actually see. Two important facts:

1. Being in the graph doesn’t guarantee a panel. Google can “know about” an entity without ever showing a panel for it. A panel appears only when Google is confident enough.
2. You can’t summon a panel into existence. If Google already shows one for you, you may be able to claim it (verify you’re the real subject) and then suggest edits. But you can’t force a panel to appear for an entity Google hasn’t recognised yet — and not every panel is even claimable.

So what actually helps? The realistic answer is: become an entity Google can recognise and trust. That means being consistent about your name and description everywhere you appear, getting mentioned and cited by other reputable sites, and — if you’re a local business — keeping your Google Business Profile accurate. Wikipedia and Wikidata entries help a lot too, when they’re warranted.

For the deeper “how do I actually build those signals” playbook, see the entity SEOEntity SEO is the practice of helping search engines and AI systems clearly identify, classify, and trust the entities you represent — your brand, your people, your products — rather than just matching keyword strings. The goal is to be an unambiguous, well-corroborated entity in machine knowledge systems so AI can cite you with confidence. article. For the mechanics of writing the schema markup itself, see entity & identity schema\"Entity & identity schema\" is a practitioner label — not an official Google or schema.org category, and not exhaustive of identity-capable schema.org types — for the three main types that declare who or what is behind a site: Organization, LocalBusiness, and Person.. This article is about the layer above both: the graph and the panel themselves. Switch to the Advanced tab for the full version.

TL;DR — The Knowledge GraphThe Knowledge Graph is Google's database of entities — people, places, organizations, and things — and the factual relationships between them. It's separate from any single website's structured data: your schema markup is one of many possible inputs to the graph, not the graph itself. is Google’s entity-relationship database — the facts about “things” and how they connect — introduced in 2012 to do search on “things, not strings.” It’s distinct from a page’s own structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding.: your markup is one of many inputs to the graph, not the graph itself. A Knowledge Panel is the UI Google renders when its confidence in an entity clears a threshold; presence in the graph never guarantees a panel, and Google’s own docs say structured data doesn’t guarantee any feature will appear. You can claim an existing panel and suggest edits — you can’t create one — and not all panels are claimable. Real influence comes from consistent, multi-source entity corroboration, not any single lever. And the graph is increasingly discussed as part of how Google grounds entities in AI searchAI search uses large language models and retrieval-augmented generation (RAG) to synthesize an answer from multiple sources rather than returning a ranked list of links. Examples include Google AI Overviews, ChatGPT Search, and Perplexity. — an industry inference, not a mechanism Google has confirmed.

What the Knowledge Graph actually is

Google’s public descriptions explain purpose and data sources but do not expose a complete inclusion or ranking formula. Evidence for this claim Official or primary documentation supporting the adjacent article claim, with scope limited to the source's published description. Scope: No ranking guarantee or undisclosed system mechanics are inferred beyond the cited source. Confidence: high · Verified: 2026-07-14 Google: Introducing the Knowledge Graph Claiming a panel permits feedback and verification; it does not give direct control over all displayed facts. Evidence for this claim Official or primary documentation supporting the adjacent article claim, with scope limited to the source's published description. Scope: No ranking guarantee or undisclosed system mechanics are inferred beyond the cited source. Confidence: high · Verified: 2026-07-14 Google: Knowledge panels

Google’s own framing, from the 2012 launch, is the cleanest definition there is: it built the Knowledge Graph as a knowledge base that lets Google understand real-world entities and their relationships to one another — “things, not strings.” That last phrase is the whole idea. Traditional information retrieval matched query strings against document strings. The Knowledge Graph is a database of entities (people, places, organizations, things) and the facts and relationships between them, so Google can reason about a thing rather than just pattern-match its name.

Scale, for context: at launch in 2012 Google said the graph contained more than 500 million objects and more than 3.5 billion facts and relationships. By its 2020 explainer, Google described it as having amassed over 500 billion facts about five billion entities. (More on the scale numbers — and the ones you shouldn’t trust — in the Myths section below.)

Knowledge Graph vs. Knowledge Panel vs. your structured data

Most competing content blurs these three. Keeping them separate is the single most useful thing on this page:

The relationship runs one direction with no guarantees: your structured data is one signal that can feed the graph, the graph is Google’s database, and the panel is a UI surface Google chooses to render from that database when confident. Google draws the graph from hundreds of sources across the web — open and licensed databases, content owners, specialized providers for health, music, sports, and TV data — not just from any one site’s markup and not just from Wikipedia.

I won’t re-litigate the entity-signal hierarchy or the sameAs/Wikidata mechanics here — the entity SEOEntity SEO is the practice of helping search engines and AI systems clearly identify, classify, and trust the entities you represent — your brand, your people, your products — rather than just matching keyword strings. The goal is to be an unambiguous, well-corroborated entity in machine knowledge systems so AI can cite you with confidence. article already covers which signals correlate with entity recognition, and entity & identity schema\"Entity & identity schema\" is a practitioner label — not an official Google or schema.org category, and not exhaustive of identity-capable schema.org types — for the three main types that declare who or what is behind a site: Organization, LocalBusiness, and Person. covers the Organization / Person / LocalBusiness markup implementation. This article stays on the graph and panel layer above both.

What triggers a Knowledge Panel

Confidence thresholds and multi-source corroboration

A panel isn’t a switch that flips the moment you exist. Google generates one when its confidence in an entity — built from corroborating facts across many independent sources — is high enough to summarize that entity directly in results. That’s why household-name brands and widely-covered people get panels and a brand new one-location business usually doesn’t: the corroboration isn’t there yet.

Why “being in the graph” doesn’t guarantee a panel

Google can have an entity in the graph and still not show a panel for it. Presence and display are two different bars. So “am I in the Knowledge Graph?” and “do I have a Knowledge Panel?” are genuinely different questions — and neither one is something you can flip on with a checkbox.

How to claim and edit your Knowledge Panel

The claim flow

If Google already shows a panel for you or your organization, you can go through a verification process to claim it. In broad strokes: find your panel in search, choose the option to claim it, and verify your identity by signing in to an official profile Google recognises as belonging to that entity (linked accounts like Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results., YouTube, or verified social profiles are the usual routes). I’m describing the flow rather than quoting Google’s help pages verbatim here — those pages render via JavaScript and I couldn’t confirm exact current wording in this pass, so check the live Knowledge Panel help docs for the precise, current steps before you rely on them.

The plain caveat worth stating up front, straight from Google’s help docs: not all knowledge panels are claimable. Claim eligibility isn’t universal.

Suggesting edits after claiming

Once you’re verified as the subject, you can suggest edits to the panel — a corrected description, a preferred image, fixed facts. These go through Google’s review; you’re suggesting changes, not editing a CMSA content management system (CMS) is software that lets users create, manage, and publish digital content — like blog posts and pages — without writing raw code. WordPress, Drupal, and Joomla are the most common open-source CMS platforms.. Google’s feedback documentation states its policy plainly: “To maintain the integrity of search results, our current policy doesn’t manually create or delete Knowledge Panels.” Panels appear and disappear automatically based on Google’s own generation process — your feedback can correct facts on an existing panel, but it can’t summon one into existence or have one removed on request. For local-business panels specifically, the business-level details — hours, contact info, address — are managed through Google Business Profile, not the general knowledge-panel claim flow.

What to do if there’s no panel yet, or it isn’t claimable

You can’t claim a panel that doesn’t exist, and you can’t force one into existence. If there’s no panel, the work is upstream: build the entity corroboration that eventually earns Google’s confidence (next section). If a panel exists but isn’t claimable, you’re limited to the general feedback route for factual errors rather than the verified-subject edit flow.

TIP Check for a matching entity record before diagnosing the panel

Search the entity name and inspect the returned type, description, and identifiers with my free Google Knowledge Graph Explorer Free

1. Search the exact official entity name, then add distinguishing context if namesakes appear.
2. Treat a no-result lookup as evidence that this search found no matching record, not proof that no Google system understands the entity.
3. Keep the Knowledge Graph lookup separate from the Knowledge Panel question; graph presence does not guarantee a panel.

Tap to enlarge

The completed Google Knowledge Graph Explorer result is labeled Not found and says Acme Analytics Test Brand is not in the returned Knowledge Graph results yet. Its playbook recommends using one canonical name, publishing an unambiguous entity home page, adding accurate Person or Organization markup with sameAs links, building corroboration in appropriate independent sources, and rechecking later. This lookup result does not prove that every Google system lacks entity understanding, and it does not predict whether a Knowledge Panel will appear.

Realistic ways to influence Knowledge Graph presence

Ranked roughly by how directly Google has confirmed each one — and hedged, because none of this is a guarantee:

1. Claim an existing panel and use “Suggest edits.” The one official, direct lever — but only available once a panel already exists and is claimable.
2. Wikipedia / Wikidata presence. Widely corroborated as influential for entity recognition, though it’s not an officially-stated Google ranking input. Earn it legitimately (notability, sourcing) — don’t fabricate it.
3. Consistent entity signals across the web. The same name, description, and authoritative references everywhere you appear. Consistency is what lets Google resolve “these mentions are all the same thing.”
4. Google Business Profile, for local entities. The managed surface for business-specific panel elements.
5. Structured data — as one contributing signal, not a dedicated lever. Google’s own Organization docs are blunt: “Google does not guarantee that features that consume structured data will show up in search results.” Markup helps Google disambiguate you; it doesn’t causally force a panel. Ahrefs’ study tracking 1,885 pages that added schema foundA 302 (\"Found\") is a temporary redirect: it forwards users to a new URL while telling search engines the original URL should stay in the index. It's a weak canonicalization signal, not the zero-equity dead end of SEO folklore. no major uplift in AI citationsAn AI citation is the visible source link an AI answer engine shows next to its generated text — the clickable reference that credits the web page it used. A citation's presence is a separate thing from whether the cited page actually supports the statement, and from being retrieved (read behind the scenes) or merely mentioned (named without a link); citation is driven more by brand mentions and being retrievable than by traditional ranking., a useful real-world check on the “add schema, get the panel” instinct.

The through-line: frame every tactic as improving eligibility odds, never as guaranteeing a panel. There is no combination of SEO that entitles you to one.

Bing’s entity graph, briefly

Bing has its own entity graph — publicly code-named Satori at its 2013 launch, though I’d treat that as historical background rather than Bing’s current official terminology. Back in 2014 Bing shipped a Knowledge Widget in Bing WebmasterMicrosoft's free portal for monitoring and improving how a site appears in Bing search — the peer to Google Search Console, plus IndexNow instant indexing, richer backlink data, and keyword volumes. Because Bing's index also feeds Microsoft Copilot, it doubles as a window into AI-search visibility. Tools that let webmasters detect and visualize entities on a page and link them to the entity graph inside Bing’s knowledge repository. I could not confirm in this pass whether that widget is still a live, supported feature in 2026 — the original 2014 announcement URL no longer resolves to that post; it now redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't. into the Bing Webmaster Blog’s current feed — so treat the widget as precedent for Bing’s entity tooling, not a current recommended action, and check Bing Webmaster ToolsThe free, first-party consoles search engines give site owners — Google Search Console and Bing Webmaster Tools — to see how their pages are crawled, indexed, and ranked, and to fix problems. The legacy name is 'webmaster tools.' directly for what’s actually live.

The honest summary: Bing’s public-facing documentation on its entity graph is much thinner than Google’s. That’s worth saying plainly rather than padding the section with secondary-source guesswork.

Why the Knowledge Graph matters for AI search

This is the part that’s changed, though it’s worth being precise about what’s confirmed and what’s inference. Google’s own Knowledge Panel documentation describes automatic panel generation from web sources and a feedback/review flow — it doesn’t publish a stated mechanism connecting Knowledge Graph presence to AI Overviews, AI Mode, or Gemini citations. What’s out there is industry inference: Ahrefs’ Knowledge Graph explainer, for example, argues the graph is no longer just a mechanism for winning a Knowledge Panel and is now part of how Google decides which brands and entities to surface in AI answers. That’s a plausible read of how groundingGrounding is anchoring an AI model's answer to source documents it retrieves at the moment you ask — not to the patterns frozen into its weights during training. Retrieval-Augmented Generation (RAG) is the most common way to do it. systems generally work, not a Google-confirmed pipeline.

I’d treat it as a reasonable hypothesis, not a mechanism: if AI answers lean on an entity database to help decide who’s real, then being an unambiguous, well-corroborated entity in that database is a reasonable thing to aim for — but nobody, including Google, has published a causal chain from “in the Knowledge Graph” to “cited in an AI Overview.” Don’t treat a Knowledge Graph Search API hit, a Knowledge Panel, and an AI citation as proof of one another — they’re separate systems with separate, and mostly undocumented, triggers. The entity work is still worth doing — see the entity SEOEntity SEO is the practice of helping search engines and AI systems clearly identify, classify, and trust the entities you represent — your brand, your people, your products — rather than just matching keyword strings. The goal is to be an unambiguous, well-corroborated entity in machine knowledge systems so AI can cite you with confidence. article for the generative-engine-optimization angle — just don’t oversell the mechanism behind it.

Common myths

• “Adding schema markup will get me a Knowledge Panel.” No. Google explicitly says structured data doesn’t guarantee any feature will show up. Schema is disambiguation help, not a panel trigger.
• “The Knowledge Graph and my schema markup are the same thing.” No. The graph is Google’s database; your schema is one of many inputs Google may or may not use.
• “The graph now has 1.6 trillion facts about 54 billion entities.” This number circulates widely in secondary SEO content, but I couldn’t trace it to any official, datable Google statement. The last figure Google itself stated on the record — in its 2020 post — was over 500 billion facts about five billion entities. Google hasn’t published an updated figure since then as far as I can verify, so don’t repeat the bigger number as fact.
• “Knowledge Vault is the current version of the Knowledge Graph.” No. Knowledge Vault was a distinct 2014 Google Research project reported by New Scientist; Google told Search Engine Land at the time it was a research report, not an active Google service. Don’t conflate the two.
• “You can pay to guarantee a panel.” No. The official claim/verification path involves no payment. Paid “knowledge panel management” services exist, but they can’t guarantee an outcome Google itself won’t guarantee.
• “Every brand is eligible for a panel with enough SEO.” No. Google’s own docs say not all panels are even claimable — eligibility and confidence thresholds are gatekept by Google’s criteria, not fully in your control.

FAQs

What’s the difference between the Knowledge Graph and a Knowledge Panel? The graph is Google’s database of entities and facts; the panel is the visible summary Google renders from it when confident. The graph can know about you without ever showing a panel.

How do I get a Google Knowledge Panel? You can’t force one. Build genuine, consistent, multi-source entity corroboration; if Google grows confident enough it may generate a panel, which you can then claim and edit. There’s no guaranteed path.

Why doesn’t my business have a Knowledge Panel? Most likely Google’s confidence in your entity hasn’t cleared its threshold yet — not enough independent corroboration across the web. It’s not something a single schema change fixes.

Does structured data guarantee a Knowledge Panel? No — Google’s Organization docs state it doesn’t guarantee that features consuming structured data will appear.

Does Bing have a Knowledge Graph? Yes, its own entity graph, but Bing documents it far more thinly than Google documents theirs.

Do I need a Wikipedia page to get a Knowledge Panel? Not strictly, but a legitimate Wikipedia/Wikidata presence is a widely-corroborated help for entity recognition. Don’t fabricate one to game it.

How big is Google’s Knowledge Graph? The last officially confirmed figure is from Google’s 2020 post: over 500 billion facts about five billion entities. Treat newer, larger numbers you see quoted as unverified unless they trace to a Google source.

AI summary

A condensed take on the Advanced version:

• The Knowledge GraphThe Knowledge Graph is Google's database of entities — people, places, organizations, and things — and the factual relationships between them. It's separate from any single website's structured data: your schema markup is one of many possible inputs to the graph, not the graph itself. = Google’s database of entities (people, places, organizations, things) and the facts/relationships between them. Introduced in 2012 to do search on “things, not strings.”
• It is NOT your structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding.. Your schema markupSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. is one of many possible inputs to the graph; the graph lives at Google’s level, built from hundreds of sources.
• A Knowledge Panel is the UI, not the database. Google renders one only when its confidence in an entity clears a threshold. Being in the graph never guarantees a panel.
• Nothing guarantees a panel. Google’s Organization docs say structured data doesn’t guarantee any feature will appear in results.
• Claim ≠ create. You can claim an existing panel and suggest edits; you can’t summon one into existence, and not all panels are claimable.
• Realistic influence: consistent multi-source entity corroboration, Wikipedia/Wikidata presence, Google Business Profile for local, structured data as a contributing (not causal) signal — never a single guaranteed lever.
• Scale to trust: last official figure is 2020’s “over 500 billion facts about five billion entities.” The circulating “1.6 trillion facts / 54 billion entities” number has no traceable official source — don’t repeat it.
• AI searchAI search uses large language models and retrieval-augmented generation (RAG) to synthesize an answer from multiple sources rather than returning a ranked list of links. Examples include Google AI Overviews, ChatGPT Search, and Perplexity. tie-in: industry analysis argues the graph is increasingly an input to how AI answers ground and verify entities — Google hasn’t confirmed this mechanism, so treat it as a reasonable inference, not a documented rule. Being an unambiguous entity is still a reasonable thing to aim for either way.

Official documentation

Primary-source documentation and posts from Google and Bing.

Google

• Introducing the Knowledge Graph: things, not strings — Amit Singhal’s 2012 launch post; the origin definition and launch-scale figures.
• Google’s Knowledge Graph and Knowledge Panels — Danny Sullivan’s 2020 explainer; sourcing (“hundreds of sources”), the 2020 scale figure, and how subjects claim panels.
• Get verified on Google / claim a knowledge panel — the claim + verification flow (JS-rendered; confirm current steps in-browser).
• Submit feedback on content about you — the suggest-edits / feedback flow after verification.
• Organization structured data — includes the explicit “does not guarantee” language and how Organization properties can influence the logo shown in your panel.
• Google Knowledge Graph Search API — developer reference for querying entities in the graph.

Bing / Microsoft

• Announcing Bing Knowledge Widget & App Linking — the 2014 entity-widget announcement (historical; verify current status).
• Bing Knowledge Comes to Webmaster Tools — the Webmaster ToolsThe free, first-party consoles search engines give site owners — Google Search Console and Bing Webmaster Tools — to see how their pages are crawled, indexed, and ranked, and to fix problems. The legacy name is 'webmaster tools.' rollout of the Knowledge Widget (2014).

Quotes from the source

On-the-record statements, with deep links that jump to the quoted passage where the source page supports it.

Google — the origin definition

• “We’ve built the Knowledge GraphThe Knowledge Graph is Google's database of entities — people, places, organizations, and things — and the factual relationships between them. It's separate from any single website's structured data: your schema markup is one of many possible inputs to the graph, not the graph itself., a knowledge base that enables Google to understand real-world entities and their relationships to one another: things, not strings.” — Amit Singhal, SVP Engineering, Official Google Blog, May 16, 2012. Jump to quote

Google — scale (2020, most recent official figure)

• The Knowledge Graph “has amassed over 500 billion facts about five billion entities.” — Danny Sullivan, “Google’s Knowledge Graph and Knowledge Panels,” Google Blog, May 20, 2020. Read the post Sourced via a summarized fetch of the 2020 post, not a raw-HTML verified substring in this pass — confirm the exact wording against the live page before treating it as final. Note also this is the last officially stated figure; larger numbers circulating on secondary sites are unverified.

Google — the explicit non-guarantee

• “Google does not guarantee that features that consume structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. will show up in search results.” — Google Search Central, Organization structured data documentation. Jump to quote Verify the exact anchor textAnchor text is the visible, clickable text of a hyperlink. It tells readers what they'll find on the other end and gives search engines context about the linked page. against the live page before treating as final.

Google — not all panels are claimable

• Google’s Knowledge Panel help documentation states plainly that not all knowledge panels are claimable — claim eligibility isn’t universal. Paraphrased, not quoted verbatim: the Knowledge Panel help pages render via JavaScript and their exact current wording couldn’t be confirmed in this pass. Verify in-browser before quoting.

”Should I worry about a Knowledge Panel?” — a decision tree

Work top to bottom.

1. Does Google already show a panel when you search your entity?

• Yes → go to step 2.
• No → you can’t claim what doesn’t exist. Skip to step 4 (build corroboration).

2. Is there a “Claim this knowledge panel” option on it?

• Yes → claim it (verify via a linked official profile), then use “Suggest edits” for any wrong facts or a preferred image. Done.
• No → the panel isn’t claimable. Use the general feedback route to flag factual errors; you can’t do verified-subject edits. Then also do step 4.

3. Are you a local business specifically?

• Yes → manage hours/contact/address through Google Business Profile, not the general panel claim flow. (Do steps 2 and 4 too.)
• No → continue.

4. No panel, or want to strengthen the entity? Build corroboration:

• Consistent name + description everywhere you appear.
• Earned mentions/citations from reputable independent sources.
• Legitimate Wikipedia/Wikidata presence if warranted (don’t fabricate).
• Complete, accurate Organization/Person structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. (one signal, not a guarantee).
• Then wait — a panel appears only once Google’s confidence clears its threshold, if it ever does. There’s no “submit” button.

The rule underneath all of it: you influence eligibility, never the outcome.

Knowledge-entity checklist

A pass to confirm you’ve done what’s actually in your control:

• Searched your entity to see whether a panel already exists.
• If a panel exists and is claimable, claimed it via a linked official profile.
• Used “Suggest edits” to fix wrong facts / set a preferred image (verified account only).
• For local: Google Business Profile is claimed and its hours/contact/address are accurate.
• Name and description are consistent across your site, socials, and third-party listings.
• Authoritative references (Wikipedia/Wikidata, verified profiles, registries) exist where legitimately warranted — and agree with each other.
• Organization / Person structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. is complete and accurate (treated as one signal, not a panel trigger).
• You’re not relying on any “guaranteed panel” service or expecting schema alone to produce a panel.
• You’re not repeating unverified scale numbers or conflating Knowledge Vault with the live graph in your own content.

The mental models

1. Three layers, one direction. Your structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. → (maybe) feeds → the Knowledge GraphThe Knowledge Graph is Google's database of entities — people, places, organizations, and things — and the factual relationships between them. It's separate from any single website's structured data: your schema markup is one of many possible inputs to the graph, not the graph itself. → (maybe) renders → a Knowledge Panel. Every arrow is “maybe,” and you only fully control the first box. Keep the three layers separate and most Knowledge-Graph confusion disappears.

2. Presence vs. display. “Is my entity in the graph?” and “do I have a panel?” are different questions. Google can know about you without showing anything. Display requires a confidence threshold that presence alone doesn’t meet.

3. Eligibility, not entitlement. You influence the odds of a panel; you’re never entitled to one. Reframe every tactic as “does this make Google more confident about my entity?” — not “does this force a panel?”

4. Claim vs. create. You can only claim (and then edit) a panel that already exists. Nothing you do creates a panel directly; creation is Google’s call, downstream of corroboration.

5. Corroboration over declaration. Declaring facts about yourself (on your own site, in your own markup) is weak on its own. Independent sources agreeing about you is what builds Google’s confidence. This is the same principle that governs entity SEOEntity SEO is the practice of helping search engines and AI systems clearly identify, classify, and trust the entities you represent — your brand, your people, your products — rather than just matching keyword strings. The goal is to be an unambiguous, well-corroborated entity in machine knowledge systems so AI can cite you with confidence. and AI-search groundingGrounding is anchoring an AI model's answer to source documents it retrieves at the moment you ask — not to the patterns frozen into its weights during training. Retrieval-Augmented Generation (RAG) is the most common way to do it..

Knowledge Graph — cheat sheet

Graph vs. Panel vs. your markup

What you can vs. can’t do

• Can: claim an existing, claimable panel; suggest edits once verified; manage local details via Google Business Profile; build entity corroboration.
• Can’t: create a panel on demand; guarantee a panel with schema; pay Google for one; edit a panel you haven’t verified for.

Scale numbers

• 2012 launch: 500M+ objects, 3.5B+ facts/relationships.
• 2020 (last official): 500B+ facts, 5B entities.
• “1.6 trillion facts / 54 billion entities”: unverified — no traceable official source. Don’t repeat as fact.

Don’t confuse

• Knowledge Graph ≠ your structured data.
• Knowledge Graph ≠ Knowledge Panel (database vs. UI).
• Knowledge Graph ≠ Knowledge Vault (Vault was a 2014 research report, not the live service).

Knowledge Graph mistakes to avoid

Treating a Knowledge Panel as the Knowledge Graph

The panel is one visible search feature generated from entity understanding. An entity can exist in the graph without a panel, so do not use panel presence as the sole test.

Trying to create notability with schema alone

Structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. can clarify first-party claims but cannot substitute for independent, reliable corroboration. Fix source consistency and real-world evidence first.

Claiming or editing the wrong namesake

Similar names are common. Verify the official URL, category, people, location, and identifiers before requesting control or suggesting changes.

Chasing every possible property

More markup is not automatically clearer. Prioritize accurate identity-defining facts and relationships that the page and external sources support.

Prompts for Knowledge Graph investigations

Reconcile these search-result observations, Knowledge Graph records, official pages,
and third-party profiles. Identify which records refer to the same entity, which are
possible namesakes, the attributes that disambiguate them, and all conflicting facts.
Return a source-backed correction plan. Do not merge entities or assert a panel cause
without evidence.

[paste records, URLs, and observations]

Review this proposed Knowledge Panel edit. Separate facts supported by the official
site, facts corroborated independently, and unsupported claims. Rewrite the evidence
note concisely and list the exact source needed for every unsupported change. Do not
use schema markup itself as independent proof.

[paste proposed edit and sources]

Tools for Knowledge Graph SEO

• Google Knowledge Graph Explorer: Search entity records and identifiers to investigate resolution and namesake collisions.
• Entity Coverage Analyzer: Check whether a page clearly states the people, organizations, products, and relationships that define the subject.
• Schema Markup Validator: Validate entity markup after confirming each stated fact and identifier.
• Google Search and Knowledge Panels: Observe panels, sources, and ambiguity for a controlled query set; claim eligible panels through the interface Google provides.
• Wikidata Query Service and item histories: Inspect identifiers, statements, and provenance without assuming Wikidata alone determines Google’s output.

Resources worth your time

My related writing

• The Beginner’s Guide to Technical SEO — where entities and structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. fit in the bigger picture.

My speaking

• How Search Works (SlideShare) — my walkthrough of crawlingCrawling is how search engines use automated bots (like Googlebot and Bingbot) to discover URLs and download pages. A page has to be crawlable to be indexed, but crawling on its own isn't a ranking factor., renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM., indexingStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed., and ranking, with the entity-understanding context around it. (Standing disclaimer applies: this is my understanding of these systems, not going to be 100% complete or accurate.)

From around the industry

• What Is Entity SEO? A Simple Guide (Ahrefs, Si Quan Ong) — the entity-signal side of this story; a colleague’s guide (“entity SEOEntity SEO is the practice of helping search engines and AI systems clearly identify, classify, and trust the entities you represent — your brand, your people, your products — rather than just matching keyword strings. The goal is to be an unambiguous, well-corroborated entity in machine knowledge systems so AI can cite you with confidence. is just SEO”) that quotes Patrick.
• Google’s Knowledge Graph Explained: How It Influences SEO & AI Search (Ahrefs, Despina Gavoyannis, reviewed by Ryan Law) — good on the AI-search-groundingGrounding is anchoring an AI model's answer to source documents it retrieves at the moment you ask — not to the patterns frozen into its weights during training. Retrieval-Augmented Generation (RAG) is the most common way to do it. angle.
• We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved (Ahrefs, Louise Linehan & Xibeijia Guan) — the empirical check on “add schema, get the win.”
• What is the Knowledge Graph? How it affects SEO and visibility (Search Engine Land, Curtis Weyant) — a structural overview of the graph and panels.
• When and how to use knowledge graphs and entities for SEO (Search Engine Land, Paul DeMott) — one practitioner’s framing of practical entity levers.
• How schema markup fits into AI search — without the hype (Search Engine Land, Aimee Jurenka) — a level-headed take on schema’s real role in AI searchAI search uses large language models and retrieval-augmented generation (RAG) to synthesize an answer from multiple sources rather than returning a ranked list of links. Examples include Google AI Overviews, ChatGPT Search, and Perplexity..
• Knowledge Graph (Google) (Wikipedia) — background and the Knowledge Vault clarification.

Test yourself: Knowledge Graph SEO

Five quick questions on the graph, the panel, and what you can actually influence. Pick an answer for each, then check.

Knowledge Graph

The Knowledge Graph is Google's database of entities — people, places, organizations, and things — and the factual relationships between them. It's separate from any single website's structured data: your schema markup is one of many possible inputs to the graph, not the graph itself.

Related: Entity SEO, Entity & Identity Schema, Structured Data

Knowledge Graph

The Knowledge Graph is Google’s own database of entities and the relationships between them. Google introduced it in 2012 to move search from matching text strings toward understanding real-world things — the framing it used was “things, not strings.” An entity is any distinctly identifiable thing (a person, a company, a place, a book, a concept); the graph stores facts about each one and the connections between them, so Google can answer “who is the CEO of this company” or “what films did this director make” without re-reading the open web every time.

The single most important distinction: the Knowledge Graph is not the same as your website’s structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding.. Structured data (schema markupSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor.) is one signal that can feed information into the graph, but the graph lives at Google’s level and is built from hundreds of sources — open and licensed databases, Wikipedia and Wikidata, content owners, and Google’s own confidence scoring — not something any one site owns or controls.

A Knowledge Panel is the visible UI Google generates from the graph when it’s confident enough about an entity to summarize it directly in search results. The graph can “know about” an entity without ever showing a panel; a panel requires Google’s confidence threshold to be met, not just presence in the graph. Nothing guarantees a panel — Google’s own docs state it doesn’t guarantee that features which consume structured data will appear in search.

Related: Entity SEO, Entity & Identity Schema, Structured Data

Changelog