Knowledge base / Structured data
Entity pages: how a machine works out that your business is a real thing.
Search engines stopped matching strings and started resolving things. If nothing on your site states plainly what your business is, something else will decide for you.
Key takeaways
- What is an entity? A thing a search engine recognises as distinct: a business, a person, a place, a service. Your business is an entity whether or not your site helps anyone resolve which one you are.
- What is an entity page? A page whose job is to state facts about one entity plainly: an About page for the company, a bio page for a named person, a service page for one service. It exists to be the reference, not to sell.
- Why does it matter for AI answers? A model asked who to call has to resolve your name to a business with a location and a service. When your site never says it in one place, the model fills the gap from whatever else it has seen.
- What is the minimum? Three pages: an About page with the real business facts, a bio for the human who signs the work, and one page per service you actually sell.
What changed, and why it matters now
Search used to match text. You typed a phrase, the engine found pages containing it. Modern search resolves entities instead: it decides which business, person or place you mean, and then finds what it knows about that thing.
Answer engines inherited this and made it sharper. When someone asks an assistant who to call for a service in a city, the model has to resolve a set of candidate businesses and say something true about each. Everything it says about you comes from what it managed to resolve. If your site never states your service area in plain words, the model uses whatever else it has seen, which might be a six-year-old directory listing.
Which pages do the entity work?
| Page | The entity it defines | What it has to state plainly |
|---|---|---|
| About | The business. | Legal name, trading name, year founded, where you operate, what you do, how to reach you. |
| Team or author bio | The people. | A real name, a role, credentials, and the pages they are responsible for. |
| One page per service | Each service. | What the service is, who it is for, where it is offered, and what it costs or how it is priced. |
| Contact | The place. | Address, phone and hours in text, never only in an image or a map embed. |
None of these are glamorous pages and all of them are usually thin, because they get judged on traffic. They are not traffic pages. They are reference pages, and they are what the rest of your site is resolved against.
The test: would a stranger get it right?
Open your About page and read it as though you know nothing about the business. Can you answer these five, from the page alone, without inferring?
- What is this business called, exactly, including whether it is Inc or LLC?
- What does it sell, in plain words rather than a slogan?
- Where does it operate, named rather than implied as “the local area”?
- Who is responsible for the work, and what qualifies them?
- How long has it existed?
Most About pages we read fail three of the five, usually the last three. Every gap is a place where a model has to guess, and a guess is where a wrong answer about your business comes from.
When your own sources disagree
The most common entity problem is not absence. It is contradiction. The website says one trading name, the Business Profile says another, an old directory lists a phone number from two offices ago, and the footer still carries a suite number you moved out of.
Every one of those is a small thing. Together they tell a resolution system that it may be looking at more than one business, which is the worst outcome: not a wrong answer, but a hedged one, where the assistant declines to recommend anyone specific.
Fixing it is unglamorous and cheap. Pick the canonical version of your name, address and phone, write it down, and make every source match it. NAP consistency covers the method, and it is part of the AI Citation Kit we run on builds and uplifts.
Questions about entities
A knowledge panel is one visible outcome, and it is Google's to grant rather than yours to build. Entity work is about making the facts about your business consistent and machine-readable. Sometimes a panel follows. The work is worth doing regardless, because AI answers draw on the same resolution problem.
More than its traffic suggests. It is the page that states what the business is, when it started, where it operates and who runs it, and it is usually the page an engine resolves you from. Judging it on pageviews is how it ends up three sentences long.
That is a real tension and worth naming. An entity needs a human somewhere for author and Person markup to mean anything. The usual middle ground is an editorial team page with real credentials and no personal details, which is what we do on our own About page and what Quiet Client protects for our clients.
One per service you would actually quote for. Not one per keyword variation. Three pages about drain cleaning split the signal for drain cleaning and give an engine three candidates where it needed one.
No, and chasing one is a poor use of time for most service businesses. What helps is being described consistently wherever you already appear: your Business Profile, your directory listings, your social profiles. See NAP consistency.
Schema is how you state the entity facts in machine-readable form, and the entity pages are where those facts live in human-readable form. They have to agree. The schema guide covers the markup side.
Check whether a machine can tell who you are.
Entity clarity is point two of the free AI Visibility Score: whether your business name, city and phone number are stated where a machine can read them.
Get your AI Visibility Score