Entity SEO – what an entity is and why it decides how Google and AI understand you

Definition

An entity is an identifiable concept, object or idea that Google and AI systems can understand semantically and relate to other information – beyond individual words or phrases. Entity SEO is the work of establishing your company as such an entity.

An entity is a distinct, identifiable thing: a company, a person, a place, a product, an event or a concept. What matters is that it can be identified unambiguously and told apart from everything else. For Google and AI systems, entities are the very basis for understanding the world – and therefore for understanding your company.

From strings to things

In 2012 Google launched its Knowledge Graph with a phrase that has shaped SEO ever since: things, not strings. According to Google, the search engine would stop treating a search as a string of characters to match, and instead understand it as a real thing with meaning and context.

The difference is fundamental. An older search engine that came across the word Apple saw five letters to look for on web pages. An entity-based search engine instead asks itself: does the user mean the company, the fruit or the record label – and what do I know about that thing?

What counts as an entity?

An entity does not have to be physical. Google describes it as something that is unique, well defined and distinguishable. In practice, the most common types are:

• Organisations – companies, public authorities, brands.
• People – founders, experts, public figures.
• Places – cities, addresses, regions.
• Products and services – a specific model or a subscription.
• Concepts and topics – abstract ideas such as conversion optimisation or headless CMS.
• Events – launches, conferences, dates.

Your company is an entity. So are your founder, your services and the topics you want to be known for. They are connected – and it is precisely the connections that give them meaning.

How search engines and AI use entities

Google collects entities and their relationships in an enormous database – the Knowledge Graph. There, your company is not a line of text but a node connected to other nodes: your services, your industry, your location, the people behind it.

Those relationships explain why a page can rank for queries whose exact wording does not even appear on it. If Google understands that your company is an entity within a topic, it can show you for related searches because of the connections – not because of exact word matching.

AI models work in a similar way, in two steps. First entity recognition: recognising that a word in a text refers to a real thing. Then disambiguation: deciding from the context which thing it is.

Disambiguation – when your name is not unique

Disambiguation is where many brands lose visibility without knowing it. If your company name is shared with another, more established entity – a well-known product, a place, a common word – AI and search engines may read content about you as being about something else.

The solution is not to change your name, but to give the systems enough unambiguous signals: consistent information, clear connections between your channels and structured data that explicitly says the name refers to this particular organisation. In practice, that work is what building an entity profile involves.

Entities or keywords – what is the difference?

They are easy to mix up. A keyword is a string of text. An entity is the thing behind the string.

"Web agency Stockholm" is a search term. SEOS Design is an entity – a company with a location, services, people and a history. It calls for a very different way of thinking: instead of chasing individual phrases, you build a clear, coherent picture of what your company is and how it relates to the world around it. The keywords then follow.

Why entities decide your visibility in AI answers

In traditional search you could get a long way on word matching and links. In AI-generated search that is not enough. An AI chooses what it highlights based on how well it understands – and trusts – an entity.

If a model does not know for certain what your company is, what you do and that different mentions are actually about you, you often drop out. Not because the information was wrong, but because it was unclear. And if the AI lacks information about you, it also tends to give an evasive or negative answer rather than saying "I don't know" – and that costs you business directly.

How a company becomes a recognised entity

Becoming a clear entity is work, not a setting you switch on. In short, it takes:

• Consistent information about the company everywhere it is mentioned.
• Structured data that explicitly connects your channels.
• External, credible mentions that confirm you exist and what you do.

The overall picture that search engines and AI build of you is called an entity profile, and that is what you are really working on over time. We cover how to appear in AI answers in AI SEO 2026 and AI Overviews and GEO. If you want to review how clearly your company comes across as an entity, get in touch.