AI SEO & GEO: Rank in Google, Get Cited by AI

Search is splitting in two. The classic job - ranking a page in Google - still pays the bills. But a growing share of research now ends inside an AI answer: ChatGPT, Perplexity, AI Overviews. Those systems don’t rank pages. They quote passages. Winning there is a different job with different rules, and it has a name now: generative engine optimization, or GEO.

The two jobs overlap more than they conflict. Content that states facts plainly, cites real sources and answers the question early tends to do well in both. But the differences matter, and they’re exactly where most SEO advice written before 2024 quietly went stale.

What changes, in practice

Three shifts drive everything in this category.

The passage beats the page. Answer engines chunk your content and retrieve the pieces. A brilliant page whose answer is scattered across four paragraphs loses to a mediocre page that states it in one. Structure has become a ranking factor in a much more literal way than “use h2s” ever meant.

The citation replaces the click. Your content can influence a buyer who never visits your site. That’s a real win that your analytics can’t see, which breaks measurement in ways worth understanding before your traffic reports start lying to you.

Freshness and authorship are load-bearing. Answer engines need an entity to attribute claims to and a date to trust them by. Anonymous, undated content is a weak citation candidate no matter how good it is.

The guides

Zero-click strategy for AI search: what still wins - the strategic layer over everything below: sorting keywords by what a click is still worth, winning citations where clicks are gone, and converting the scarce visits that remain.

How to get your content cited by AI search engines - the practical playbook: writing sections that make sense on their own, the structure signals that actually matter, and how to measure citation when no dashboard shows it.

Passage-level citability: write sections AI can quote - the construction manual one level down: the standalone test every section has to pass, the structure that marks chunk boundaries, and Google’s own word that no special markup exists.

AI Overview visibility: what Search Console now shows - Google’s generative AI report reached every site on 31 August 2026. What it measures, what it withholds, and how to cover the assistants it will never report on.

Bing Copilot citations: how to get into Copilot answers - the AI Performance report in Bing Webmaster Tools shows citations and grounding queries no other platform reports, and the monthly loop that uses them.

Perplexity citations: content SEO for AI agents - the assistant built around the citation: how its two crawlers reach your content, the WAF and robots.txt settings that silently disqualify sites, and why it’s the best canary for citation work.

ChatGPT search optimization: get found in AI answers - the biggest answer surface and the least transparent one: the three-crawler distinction that decides whether you can appear, what its citation data says gets cited, and what to skip.

Measuring AI referral traffic: GA4 setup and attribution - the 30-minute custom channel that pulls ChatGPT, Perplexity and Copilot visits out of the referral haystack, and how to read the small numbers without fooling yourself.

How to use ChatGPT for keyword research - what the model is good at (expansion, intent, clustering by meaning), what it fabricates (every volume and difficulty score), six prompts, and the validation step against real data.

Google’s AI content policy: what it actually says - the two clauses the whole policy reduces to, the scaled content abuse line the March 2024 crackdown enforced, and why disclosure is recommended but not required.

E-E-A-T for AI content: proving expertise when you use AI - what the rater guidelines actually say about AI content and fake author profiles, the position-1 data on experience, and where to make yours visible in a model-assisted post.

AI content plus human editing: the formula that ranks - what the million-page studies show about edited versus raw AI content, the five-pass edit that closes the gap, and the survey finding that AI editing AI performs worst of all.

Generating schema markup with ChatGPT, safely - the one genuinely good model-plus-marketer pairing in technical SEO: the prompt that forbids invented properties, the two validators that judge the output, and the honest list of what schema won’t do.

Technical SEO audit with AI: checks you can run yourself - six checks that need a browser and a paste, not a crawler license: robots.txt read like Google reads it, canonicals, title hygiene at scale, and PageSpeed reports translated into tickets a developer will act on.

Image alt text at scale: AI descriptions that rank - the most neglected field in SEO fixed in an afternoon: the quality bar that separates useful alt text from generated spam, the batch workflow with a vision model, and the images that stay human-written.

The measurement side of the zero-click problem - what to do when your attribution model can’t see the AI touchpoints doing the persuading - lives in the Analytics & ROI topic.

If SEO is your job title, the AI guide for SEO specialists organises all of this by task. More guides are coming here: llms.txt in practice and how AI Overviews choose their sources. For the wider picture of where search fits an AI marketing practice, start with the complete guide to AI marketing.