The knowledge-graph mindset

Why entities aren’t enough on their own

The last two lessons made the case for entity clarity — being clearly identifiable as a specific thing. But being recognisable isn’t the same as being recommended. A clear entity that isn’t connected to anything useful is still invisible when someone asks for help.

What makes you recommendable is the web of relationships around you. The topics you’re associated with. The places you operate in. The people you’ve worked with. The kinds of questions you tend to answer.

AI systems represent these relationships as a knowledge graph — a network of entities connected by labelled links. You don’t need to understand the technical details to use the idea. You just need to start thinking about your business as a node in a network rather than a stand-alone page.

What a knowledge graph actually is

A knowledge graph is a structured map of entities and the relationships between them.

A simple example. Imagine three entities: Warren Groom, Toronto, and WordPress development. On their own, they’re just three things. A knowledge graph connects them with explicit relationships: Warren Groom is based in Toronto. Warren Groom specialises in WordPress development. WordPress development is a kind of web development. Toronto is a city in Canada.

Multiply that across thousands of entities and millions of relationships, and you have something close to how an AI system understands the world.

When someone asks an AI “who’s a good WordPress developer in Toronto?”, it isn’t matching keywords. It’s traversing the graph — looking for entities of type WordPress developer, with a based in relationship to Toronto, ideally with strong trust signals attached. The answer it returns depends entirely on which relationships are visible to it.

Your job, as a website owner, is to make your relationships visible.

The relationships that matter most

You don’t need to map every possible connection. A few categories do most of the work.

Topic relationships. What subjects are you associated with? If you’re a freelance developer who works exclusively on WordPress, every part of your site should reinforce that connection — your services, your blog posts, your portfolio, your bio. Mentions of unrelated platforms dilute the link. Repeated, specific mentions of WordPress strengthen it.

Place relationships. Where do you operate? “Based in Toronto” appears in your footer, your About page, your Google Business Profile, your contact page, and any geographical schema markup you have. The more consistently you state the relationship, the more confidently an AI can use it.

Audience relationships. Who do you serve? “I work with marketing agencies” or “for B2B SaaS companies” or “for small dental practices in Ontario.” These relationships connect you to the kinds of buyers who ask AI for help finding someone like you.

Method relationships. How do you work? “Hand-coded custom themes, no page builders.” This kind of specific methodology becomes a relationship — Warren Groom prefers hand-coded development. It’s how AI systems can distinguish you from generic options.

People relationships. Who have you worked with, who recommends you, who are you connected to professionally? Named clients, testimonials with attribution, and links to collaborators all build the social graph around your business.

None of these are revelatory in isolation. They’re things any well-written site mentions. The shift is doing them deliberately — knowing that each mention is reinforcing a specific relationship in a knowledge graph.

How to shape your knowledge graph deliberately

You can take a fairly mechanical approach to this. It works.

Pick five to ten relationships you want your business to be known for. Write them out as plain sentences: “I am a freelance WordPress developer based in Toronto. I work primarily with marketing and PR agencies. I specialise in custom themes and white-label delivery. I have been doing this since 2008. I do not use page builders.”

Then ask yourself, for each relationship: is this stated clearly on my site? Is it stated in more than one place? Is it stated the same way on my LinkedIn, my directory listings, my Google Business Profile? Does my actual published content reflect it?

The relationships that pass all four checks are the ones AI systems will pick up and use. The ones that don’t are the ones you have an opportunity to strengthen.

This is the most underused lever in GEO, and it costs nothing to use.

A small warning

There’s a temptation, once you understand this, to try to associate your business with as many topics as possible. Resist it.

A knowledge graph rewards specificity. A business that’s clearly connected to three topics gets recommended for those three topics. A business that’s loosely connected to twenty topics gets recommended for none of them. Spreading your associations thin doesn’t make you more discoverable — it makes you less.

Pick the relationships that are actually true, that you actually want to be known for, and that you can actually reinforce consistently. Then reinforce them.

A useful mindset

You’re not building a website. You’re building a place in a network — and the relationships you reinforce decide where in the network you end up.

The websites that win in the next few years won’t be the ones with the most content. They’ll be the ones whose relationships are clearest.

Coming up in the next module: Content structured for citation. We move from how AI systems understand who you are to how they use what you’ve written. The next four lessons cover the specific patterns — definitional sentences, question-first structure, self-contained answers — that turn good content into citable content.