Entity-based SEO for B2B websites

What is entity-based SEO?

It is the practice of building your site around entities, the real things people search for, rather than around isolated keyword strings. An entity can be a company, a product, a problem, an industry, a standard or a buyer role.

Google described this shift when it launched its Knowledge Graph as a move to "things, not strings." At launch it said the graph held more than 500 million objects and more than 3.5 billion facts about and relations between them.^[1] Google's Knowledge Graph Search API describes the graph today as having "millions of entries that describe real-world entities like people, places, and things," typed with schema.org vocabulary.^[2]

Why does it matter for B2B?

B2B buying is full of connected entities. One buying group may research a contract management product, ask whether it meets a security standard and check whether it integrates with its ERP system. Those are linked questions, not separate keywords.

An entity model helps in three ways:

• Clarity for machines. Natural language tools already pick entities out of text. Google Cloud's entity analysis, for example, "inspects the given text for known entities" and returns their names and types.^[3] Pages that name things plainly and consistently are easier to read this way.
• Clarity for buyers. One page per thing, linked along real relations, is also easier for people to follow.
• Less content sprawl. A shared list of entities tells writers what exists, what is missing and what should link to what.

How is it different from a keyword list?

A keyword list records strings and volumes. A taxonomy sorts pages into a tree. An ontology adds the things themselves and how they relate, so "data clean room" is not only a phrase but an entity with a purpose, typical buyers, required inputs and alternatives. Each relation suggests a page or a link.

How do you identify your core entities?

Start small. List the entities that define your offer:

• What you sell, and the product or service categories it belongs to.
• The problems those offers solve.
• The industries and buyer roles you serve.
• The standards, platforms and integrations buyers ask about.

Check the list against evidence: Search Console queries, sales call notes and support tickets show the words buyers really use. Model the entities that drive qualified demand first and expand later.

How do you map attributes and relations?

For each entity, write down a short definition, the questions buyers ask about it, common objections, related standards and alternatives. Then name its relations with plain verbs: is an alternative to, requires, integrates with, solves, used by.

That relation map becomes your internal linking plan. A product page links to the problems it solves and the platforms it integrates with; each problem page links back. Our guide to calculator landing pages for B2B SEO shows one way to turn a single relation into a useful page.

How does structured data support entities?

It states explicitly what a page is about. Google calls structured data a way of "providing explicit clues about the meaning of a page," and recommends JSON-LD where your setup allows it.^[4] Useful pieces for an entity model:

• Organization markup on your home or about page. Google says it can help "disambiguate your organization in search results," and its sameAs property points to other pages about your organization, such as official profiles.^[5]
• The about property, defined by schema.org as "the subject matter of an object," for the main entity of an article.^[6]
• The mentions property for entities an article refers to but "is not necessarily about."^[7]

Keep the markup honest. Google does not guarantee structured data will show in results, and its guidelines say not to mark up content that is not visible to readers.^[8]

What has changed in markup?

Two rich results the original advice leaned on are gone from Google. Google removed its How-to structured data documentation in September 2023 because that rich result was no longer shown, and the FAQ rich result stopped appearing in Google Search from May 7, 2026.^[9] The schema.org types still describe content, but they no longer earn those special listings in Google. Put the effort into clear pages and Organization, Article about and mentions markup instead.

Does it help with AI search?

Indirectly. Google says there are no additional requirements or special optimizations needed to appear in AI Overviews or AI Mode, that normal SEO best practices still apply, and that structured data should match the visible text.^[10] Clear entities, consistent facts and answer-first pages help because they make content easier to understand and quote. Our guide to generative engine optimization covers AI answers in more depth, and zero-click search strategy covers visibility when no click follows.

What mistakes should you avoid?

• Overbuilding first. Do not model thousands of entities before testing a few.
• Ignoring how buyers search. An entity matters only if buyers look for it.
• Leaving out product and sales teams. They know the real relations between offers, problems and integrations.
• Treating it as a one-off. Review the model when products, markets or queries change.

Our SEO service covers entity mapping and structured data as part of our digital marketing work.