What is Generative Engine Optimisation (GEO)?
GEO is the practice of making your business findable, understandable and citable by AI answer engines - ChatGPT, Perplexity, Claude, Google AI Overviews. It overlaps SEO heavily, then adds signals classic SEO never measured.
Why a new term at all
Search results are links you compete to appear among; AI answers are synthesised prose that names a handful of sources - or none. Being technically indexed is no longer the finish line: the engine has to understand you well enough to describe you, and trust you enough to cite you. GEO is the work of earning both.
Where GEO and SEO overlap
Crawlability, site speed, quality content, authority and links - the fundamentals feed both. A site with clean technical SEO starts ahead on GEO. Nothing in GEO licenses you to skip the basics; most GEO failures are SEO failures wearing a new name.
What GEO adds
Machine comprehension signals: content in raw HTML (most AI crawlers do not run JavaScript), structured data that states facts about your organisation and offers, llms.txt, explicit AI-crawler permissions, quotable self-contained passages, visible authorship and freshness, and a consistent entity story across your site and profiles. These rarely appeared on an SEO checklist before 2024.
What actually gets cited
Answer engines favour content that is specific, verifiable and liftable: definitions, numbered steps, comparison tables, FAQs, statistics with sources. Vague brand copy gets summarised into nothing; a crisp 40-word answer to a real question gets quoted nearly verbatim. Structure is strategy.
How to start this week
Baseline what assistants currently say about you, fix raw-HTML readability and crawler access, add Organization and FAQ schema, publish llms.txt, and restructure your key pages around real questions with direct answers. Then re-test monthly. AuditHQ’s AI Visibility suite automates the measurement half of that loop - it is one of the nine suites in the free scan.
Frequently asked questions
Is GEO just SEO rebranded?
About seventy percent of the work overlaps; the remaining thirty percent (llms.txt, AI-crawler policy, schema depth aimed at comprehension, citation-friendly structure) is genuinely new. The mistake is treating either as replacing the other.
Can you measure GEO results?
Yes, though less precisely than rank tracking: recurring test prompts across assistants, referral traffic from AI surfaces, appearance in AI Overviews, and audit scores on the underlying signals. The measurement discipline matters more than the tool.
Do AI Overviews reduce website traffic?
For pure-information queries, often yes - the answer is absorbed on the results page. For commercial and local queries, being the business the answer names is the new position one. GEO is how you compete for that placement rather than losing it silently.