Entity signals are the consistent, structured facts about your business that AI systems use to confirm you are a real organisation. They include your name, address, sector, registration details, and the way those facts appear, or fail to appear, across your website, directories, social profiles, and structured data. If those signals are weak, inconsistent, or absent, AI search tools will not recommend you. To them, you effectively do not exist.
The problem most businesses have not noticed yet
Think about how a restaurant ends up in a guidebook. The editors do not visit every restaurant in the country on the off chance. They work from consistent data: the name, the address, the type of food, the price range, the opening hours. If a restaurant cannot be found in a reliable directory, has a different name on its website than on its sign, and has never been reviewed anywhere credible, it will not appear in the guidebook no matter how good the food is.
AI systems work in exactly the same way.
ChatGPT, Perplexity, Claude, and the AI features built into search engines do not browse your website looking for the best paragraph. They draw on structured knowledge about businesses and organisations that has been confirmed across multiple sources. When someone asks one of those tools to recommend a solicitor, a plumber, or a software consultancy, the system looks for businesses whose identity is clear, consistent, and corroborated.
If your business facts are scattered, contradictory, or simply not there, you are not in the guidebook.
What entity signals actually are
An entity, in the way AI systems use the term, is a distinct, identifiable thing in the world. Your business is an entity. So is a person, a place, a product, a film. An entity has attributes: facts that define it. A business entity has a name, a legal form, a sector, a location, a set of offerings, and a history.
Entity signals are the evidence that confirms those attributes exist and are accurate.
The signals AI systems look for include:
Consistent name and contact information. Your business name, trading address, and phone number need to appear identically across your website, Google Business Profile, Companies House or equivalent registry, and any industry directories where you appear. Small discrepancies, "Ltd" on one source, no "Ltd" on another, create ambiguity. Ambiguity means lower confidence. Lower confidence means fewer recommendations.
Structured data on your website. This is code, invisible to readers, that tells machines what your business is in a formal, standardised language. Rather than leaving an AI to infer your sector from your homepage copy, structured data states it explicitly. We cover this in detail at /aeo/structured-data.
Registry and directory presence. Companies House, Google Business Profile, LinkedIn, relevant industry associations, and sector-specific directories all act as corroborating sources. The more credible sources agree on your business facts, the stronger your entity signal.
llms.txt. A relatively new file that sits on your website and tells AI systems directly what your business does, who it serves, and what it should and should not infer about you. It is the equivalent of putting a clear label on a file that might otherwise get mis-sorted. We cover this at /aeo/llms-txt.
WebMCP. A protocol that allows AI agents to interact with your business data in a structured way. Relevant for businesses whose services involve transactions, bookings, or queries that AI tools might handle on a user's behalf. We cover this at /aeo/webmcp.
Why consistency matters more than completeness
You could have your business listed on fifty directories and still have weak entity signals if the information is inconsistent. An AI system that finds your registered address listed three different ways across five sources will treat that as noise, not as confirmation. It reduces confidence in all five sources.
This is one of the most common problems we find when we run the 16-Probe Scan on a business. Not that the business is absent from the web, but that its facts are fragmented. The business exists in dozens of places, but as dozens of slightly different versions of itself.
Fixing that inconsistency is often the highest-value single action a business can take for AI visibility.
The difference between entity signals and SEO
Traditional search engine optimisation is built around keywords and links. You create content that contains the phrases people search for, and you earn links that tell Google your content is authoritative. The result is a ranking: your page appears at position four for a given query.
Entity signal optimisation is different in purpose and method.
You are not trying to rank a page. You are trying to establish a business identity that AI systems can confidently refer to when answering questions. The question might be "who are the best accountants for small manufacturers in my sector" and the answer will be a recommendation, not a list of links. To appear in that recommendation, the AI needs to be certain about what your business is and what it does.
Keywords are about what your pages say. Entity signals are about what your business is.
Both matter. They work at different levels.
What happens without entity signals
AI tools do not fail gracefully when entity signals are weak. They do not say "we could not find enough information about this business." They simply recommend businesses whose signals are clearer.
If a competitor has strong entity signals and you do not, the AI will recommend them. Not because they are better, not because they rank higher on Google, but because they are easier for the system to identify and trust.
This is happening now. Most businesses have not done this work yet. The window where doing it gives you an advantage over competitors who have not started is open, but it will not stay open indefinitely.
How we build entity signals for a business
Our process starts with the 16-Probe Scan, which scores your current entity signal strength across sixteen specific data points. That gives us a baseline: what is there, what is missing, what is inconsistent.
From there, we work through the 5-Layer Framework in the order that delivers the most impact. Entity signals sit within that framework alongside crawlability, structured data, llms.txt, and WebMCP. You can read the full framework at /frameworks/5-layer-framework.
For most businesses, the entity signal work involves:
- Auditing and correcting name, address, and contact data across all sources
- Implementing Organisation, LocalBusiness, or relevant structured data schemas on the website
- Submitting to or correcting records in key registries and directories
- Creating or updating llms.txt with accurate, machine-readable business description
- Adding schema markup to products, services, people, and FAQs where relevant
The result is a business that AI systems can find, identify, and recommend with confidence.
The baseline check
If you want to know where your business stands right now, the free visibility check at beknown.world runs the 16-Probe Scan and tells you which entity signals are present, which are missing, and which are inconsistent. It takes two minutes and produces a concrete score rather than a vague assessment.
Knowing your score is the starting point. Everything else follows from there.
Frequently asked questions
What are entity signals in AI search?
Entity signals are consistent, structured facts about your business that appear across multiple sources: your website, directories, social profiles, and structured data. AI systems read these signals to confirm that your business is real, well-defined, and trustworthy. Without clear entity signals, an AI tool has no reliable way to identify you and is unlikely to recommend you.
Why do entity signals matter for AI visibility?
AI search tools do not browse the web in real time. They work from information they have already indexed and structured. If your business facts are missing, inconsistent, or buried in prose, you are invisible to those systems regardless of how good your website looks or how high you rank on Google.
What counts as an entity signal?
Your business name, trading address, phone number, company registration, founding year, sector, products or services offered, and the consistent way these facts appear across your website, Google Business Profile, Companies House, LinkedIn, industry directories, and structured data markup. The signals need to match across all of those sources to be credible to an AI system.
How is this different from SEO?
Traditional SEO optimises for keyword relevance and link authority so a search engine ranks your pages. Entity signal optimisation tells AI systems what your business fundamentally is, not just what your pages are about. You are building a machine-readable identity, not chasing rankings. The two are related but the goal and the method are different.
How long does it take for entity signals to take effect?
Structured data is readable immediately. Directory and registry entries typically take two to eight weeks to be indexed and incorporated. Consistency across sources builds over time. Most clients see measurable improvement in AI citation rates within three months of full entity signal implementation.
Can I build entity signals myself?
Some of it, yes. You can add structured data to your website, claim your Google Business Profile, and update directories manually. The difficulty is knowing which signals matter most, catching inconsistencies across dozens of sources, and implementing the technical layers such as llms.txt and WebMCP that require development work. Most businesses find it faster to have it done properly once than to patch it incrementally.
Does this affect my Google ranking?
Entity clarity tends to improve Google performance as a side effect, because Google also uses entity signals to understand businesses. But the primary purpose here is AI search visibility: getting recommended by ChatGPT, Perplexity, Claude, and similar tools, not climbing a results page.