Generative Engine Optimisation.

Generative Engine Optimisation is the practice of improving how clearly search engines, AI assistants and answer engines understand, trust and reference your brand. It brings together SEO, structured data, entity optimisation, content strategy, digital PR, authority building and technical foundations into one approach built for how people now find information.

Answer Engine Optimisation (AEO) is part of the same picture. It focuses on structuring content so answers are clear, useful and easy for search engines and AI systems to interpret: the building blocks these tools draw on when they generate a response.

Search behaviour is changing. People increasingly ask ChatGPT, Gemini, Claude, Perplexity, Copilot and Google AI Overviews for recommendations rather than scrolling a page of links. Businesses that build authority and clarity now may improve their visibility, and the likelihood of being discovered and referenced, as AI search becomes more common.

GEO and SEO are not competing budgets. Most AI answer systems either retrieve from a live search index or were trained on a web that search engines mapped first, so a page that cannot be crawled, indexed or understood by conventional search is unlikely to be available to an AI system either. The practical implication is that the technical and authority work behind good SEO is a prerequisite for GEO rather than an alternative to it. What GEO adds is a different unit of optimisation: SEO optimises a page to occupy a position, while GEO optimises a passage to be a usable answer and a brand to be the entity worth naming.

Being referenced in an AI answer tends to follow a few conditions rather than a single technique. The content has to be retrievable, crawlable, indexed, and not locked behind scripts or logins. It has to be extractable, a claim stated plainly in a sentence or two, near a heading that matches the question, rather than buried in the middle of a 2,000-word essay. And it has to be corroborated: the same facts about your business appearing on independent sources a system has reason to trust. Specificity helps throughout: figures, dates, named locations and clear qualifications give a model something concrete to lift. Vague marketing language gives it nothing to work with.

Entity consistency is the least glamorous part of this work and one of the most useful. AI systems assemble a picture of your business from many sources, and contradictions degrade it: a different trading name on your Companies House record and your website, three variations of your address, an old phone number still live on directories, services described one way in your copy and another in your Google Business Profile. Structured data does the same job explicitly: Organization, LocalBusiness, Service, FAQPage and Article markup state in machine-readable terms what a human reader infers from layout. Neither guarantees a citation. Both remove reasons for a system to hedge, or to name a competitor it happens to understand more clearly.

Measuring AI visibility is genuinely harder than measuring rankings, and we would rather say so than present a confident dashboard. There is no rank tracker for AI answers with an agreed methodology, responses vary between users and sessions, and the systems change without notice. What we do instead is practical: maintain a set of prompts representing how your buyers would actually ask for what you sell, sample them across the main assistants on a regular schedule, and record whether the brand appears, how it is described, and which sources the answer drew on. That runs alongside referral traffic from AI tools, branded search volume, and enquiries where the prospect names an assistant as the source.

llms.txt is a proposed convention: a plain-text file at the root of a domain offering AI systems a curated map of a site's most useful content, roughly as robots.txt offers crawlers a set of rules. It is not an official standard, adoption is uneven, and no major AI provider has committed to honouring it, so we treat it as a low-cost addition rather than a lever. The related decision matters far more: your robots.txt, firewall and CDN rules determine whether AI crawlers such as GPTBot, ClaudeBot, PerplexityBot and Google-Extended can reach your site at all. We regularly find them blocked by a security plug-in default the business never consciously chose. If visibility in AI answers matters to you, that is worth checking before anything else.