12 May 2026

There is a question that increasingly keeps UK marketing teams up at night: “Why does ChatGPT recommend our competitor and not us?” It is a fair question, and the answer is almost never about who has the better product. It is about who has the better-structured digital identity. AI search engines — whether that is Google’s AI Overviews, ChatGPT Search, Perplexity, or Gemini — do not browse the web the way a human does. They rely on a web of interconnected, verified facts about entities: people, companies, places, products, and concepts. That web is called a knowledge graph, and if your brand is not in it — or is represented poorly within it — you are functionally invisible to the fastest-growing discovery channels in search. This post is the practical playbook for fixing that. You will learn exactly what a brand knowledge graph is, how AI systems use it to decide who to cite, and the specific steps you need to take right now to build one for your UK business. What is a Brand Knowledge Graph – and Why Do AI Engines Depend on It? A knowledge graph is a structured database of entities and the relationships between them. Google has maintained its own Knowledge Graph since 2012. It is the reason you can type “CEO of Apple” into Google and get a direct answer rather than a list of web pages. The graph knows that Tim Cook is a person, that he holds the role of CEO, and that Apple is a technology company headquartered in Cupertino — because those facts are linked, verified, and stored as structured relationships, not just text on a page. Modern AI search engines — including the large language models powering ChatGPT Search and Perplexity — have absorbed and extended this model. When they generate an answer about a topic, they reach first for entities they can confidently identify and verify. Brands that exist as clear, consistent, well-corroborated entities in the AI’s training data and live retrieval layer get cited. Brands that exist only as a collection of web pages — unconnected, inconsistently described, with no authoritative entity anchors — get ignored. For a UK digital marketing agency advising clients on AI visibility, this is the fundamental insight: the question is no longer just “does Google trust my website?” It is “does the AI know my brand exists as a real, verifiable entity in the world? The Five Layers of a Brand Knowledge Graph Building your brand knowledge graph is not a single task — it is a layered architecture. Each layer reinforces the others and contributes signals that AI systems use to verify and represent your brand accurately. Layer 1: Your Google Business Profile – The Entity Anchor Your Google Business Profile (GBP) is the single most important entity anchor for a UK business. It is Google’s primary mechanism for tying your brand name to a physical presence, a category, a geographic location, and a set of verified attributes. An incomplete or inconsistently maintained GBP is one of the most common reasons UK businesses fail to appear in AI-generated local and branded answers. What to do: Ensure your GBP is fully completed — not just the basics (name, address, phone number) but every available field: business description (use full sentences that define what your business is, not just what it does), primary and secondary categories, products or services listed individually, Q&A section populated with real questions your customers ask, and regular posts that demonstrate ongoing activity. Your business description should read like a factual entity definition. For example: “SEO Syrup is a London-based digital marketing agency founded in [year], specialising in search engine optimisation, paid advertising, web development and marketing automation for UK small and medium-sized businesses.” That sentence tells an AI system who you are, what you do, where you are, and who you serve — in one extractable statement. Layer 2: Schema Markup – Translating Your Website into Machine Language Schema markup is structured data embedded in your website’s code that tells search engines and AI systems what things are, not just what your pages say. For brand knowledge graph building, the most critical schema types are: Organisation schema — This should live on your homepage and include: name, url, logo, description, foundingDate, founder, address (using PostalAddress with UK-specific fields), areaServed, sameAs (linking to all your verified social and directory profiles), and contactPoint. The sameAs property is particularly powerful. It links your website’s Organisation entity to your LinkedIn company page, your Companies House record, your Wikidata entry (more on this shortly), your Crunchbase profile, and any other authoritative sources that describe your business. This creates a web of corroboration that AI systems can cross-reference to confirm your brand’s existence and attributes. Person schema — For founders, directors, and key team members, implement Person schema with name, jobTitle, worksFor (linked to your Organisation entity), sameAs (linking to their LinkedIn, Twitter/X, and any published author profiles), and a brief description. AI systems are far more likely to cite brands whose leadership team exists as verified entities in the graph. Service and Product schema — Each core service your business offers should be marked up with Service schema, including name, description, provider (linked to your Organisation), areaServed, and serviceType. This is how an AI answering “best SEO agencies in London” understands that your agency specifically offers SEO — not just digital marketing in general. A practical UK example: a Manchester-based accountancy firm implemented full Organisation schema with sameAs linking to their ICAEW member listing, their Companies House record, and their Trustpilot profile. Within three months, the firm began appearing in Perplexity AI answers for “chartered accountants in Manchester” queries — without any additional content production. The schema alone created enough entity coherence for the AI to confidently cite them. Layer 3: Wikidata and Wikipedia – The Authoritative Entity Registry Wikidata is the open knowledge base that feeds directly into Google’s Knowledge Graph, and increasingly into the training data and retrieval systems of