What is an entity in SEO, and why does it matter for AI?

An entity in SEO is a specific, identifiable thing, such as a business, a person or a place, that a search engine can tell apart from everything else. A keyword is text to be matched. An entity is the thing the text refers to. Google described this shift years ago as moving from “strings” to “things”, when it introduced the Knowledge Graph, its database of people, places and organizations and the connections between them.

Keywords versus entities

A keyword can be matched without being understood, and an entity can’t. The phrase “drain cleaning El Cajon” is a string of words that can appear on any page. A particular plumbing company based in El Cajon is an entity: it has a name, a phone number, an area it covers, services it offers and customers who have reviewed it. A page can contain the keyword a hundred times and still leave a machine unsure which company it belongs to, or whether that company exists.

What makes up a local business entity

Start with the facts that stay the same wherever your business is described. For a home-service company those are:

• The business name, written one way.
• The main phone number.
• The service area, and the address if you show one.
• The type of business and the services you offer.
• Your hours.
• The profiles that all refer to you: your website, your Google Business Profile, directory listings and social pages.
• What other people say: reviews, and mentions on local websites.

When these agree with each other across the web, a machine can treat every mention as evidence about the same company. When they disagree, it may see two companies, or one company it isn’t sure about. The name is the part that goes wrong most often, which is why your business name needs to be the same everywhere.

Why entities matter for AI answers

An AI answer to “who should I call” has to name particular businesses and say something about each one. A list of links can be vague about who is behind a page. A sentence that says “this company serves Oceanside and offers same-day repair” can’t. I can’t see inside any assistant, and the companies that make them don’t publish how local businesses are chosen. What can be observed from outside is that businesses with clear, matching facts across several sources get described accurately, and businesses with conflicting facts get left out, mixed up with a similarly named company, or described with out-of-date details. A clear entity lets machines describe you correctly, and nobody can guarantee it gets you recommended.

How structured data helps

Code added to a page can state your facts in a form a machine doesn’t have to interpret. Structured data uses the shared vocabulary from schema.org to say in labeled fields: this is a local business, this is its name, this is its phone number, these are the places it serves, and these other profiles belong to the same business. That last part, which links your website to your profiles elsewhere, is what ties the separate records together.

Structured data labels facts. It doesn’t create them. The markup has to match what a visitor can read on the page, and it has to match what your listings say. Marking up a claim that nothing else supports doesn’t make it true to a search engine. Making the markup, the page and the listings agree is the job in entity and schema work.

How to check your own business

Search for your exact business name with your city, and see what Google shows. Check whether the profile beside the results is yours and whether its phone number and hours are right, along with the category. Then ask an AI assistant what it knows about your business by name, and compare the reply with the truth. Wrong details or a blend of you and another company point to an unclear entity. So does “I couldn’t find that business”.