Skip to content

AI Search

How to get cited by ChatGPT, Claude and Google AI answers

A growing share of buying research never reaches a results page. Here is what appears to actually influence AI answers, a checklist you can run this week, and an honest account of what nobody can promise you.

By Kristian Bansil9 min read

AEO, GEO, GSO — one thing, three labels

The vocabulary has not settled, which matters mainly because it makes the field look more crowded than it is. Answer engine optimisation came first and is the term most agencies use. Generative engine optimisation arrived with the academic work on the subject and is the term you will see in papers. Generative search optimisation is the newer coinage and is gaining ground fastest in search. All three name the same job: being the source an assistant quotes.

We use AEO throughout this site because it is the oldest and least ambiguous of the three. If you searched for one of the other two, you are in the right place.

Why this matters now

A growing share of commercial research never reaches a page of blue links. Someone asks an assistant “who builds booking systems in Pampanga” and acts on the two or three names it returns. If your business isn't legible to those systems, you're not in the consideration set — and no amount of traditional ranking fixes it, because there was no results page.

This is uncomfortable for the SEO industry because the familiar levers don't map cleanly. Backlinks matter less. Being unambiguously factual and parseable matters far more.

What actually appears to influence AI answers

  1. Crawler access. If GPTBot, ClaudeBot, PerplexityBot and Google-Extended are blocked in your robots.txt, you have opted out. Many sites blocked these in 2024–25 on principle and never revisited it. Check yours today.
  2. Answerable content. Assistants quote passages that answer a question directly. A page that says “we offer competitive pricing” is unquotable. A page that says “a four-page website costs ₱75,000, fixed price, live in two weeks” can be lifted straight into an answer.
  3. Structured data. Schema.org markup removes ambiguity about what you are, where you operate and what things cost. ProfessionalService with service-area and payment data, FAQPage on question content, Article on guides, and a real price on every service.
  4. Corroboration. Facts stated consistently in several places — your site, directories, your Google Business Profile — are more likely to be treated as reliable than a claim that appears once.
  5. An llms.txt. An emerging convention: a plain-text summary at your root describing who you are, what you charge and what each page covers. Cheap to produce, and it hands an assistant exactly the summary you'd want quoted.

The practical checklist

AEO checklist: action, effort and why it matters
ActionEffortWhy it moves the needle
Unblock AI crawlers in robots.txt10 minutesBlocking them is a hard opt-out from every AI answer
Publish real pricesLow“Contact us for pricing” is unquotable; a number is citable
Add FAQPage schema to question contentLowExplicitly pairs questions with answers in machine-readable form
Write a short answer paragraph at the top of each guideLowGives the assistant a self-contained passage to lift
Ship an llms.txt1 hourYou author the summary instead of leaving it to inference
Fix NAP consistency across directoriesMediumCorroboration across sources raises confidence
Track prompt-level visibility monthlyOngoingThe only way to know whether any of it worked

How to measure something this opaque

You can't see an assistant's ranking factors, but you can build a repeatable test. Write 20–30 prompts a real customer would type — “best web developer near Clark”, “how much is a booking system in the Philippines”, “agency that builds AI chatbots Pampanga” — and run them monthly across the assistants your market uses. Record whether you were mentioned, in what position, and what was said about you. It's crude, but a trend line over six months is genuinely informative, and it catches the case where you're mentioned inaccurately, which is worse than absence.

What we do on our own site

We practise this rather than just selling it. React Web Development Services ships an llms.txt, explicitly allows the major AI crawlers, carries a full schema graph on every page, publishes every price openly, and opens each guide with a self-contained short answer. None of that guarantees a citation. All of it removes reasons to be skipped. We also measure the result and publish it: how often ChatGPT and Gemini recommend us.

Traditional SEO asks “how do we rank for this query”. AEO asks “if an assistant had to answer this question in three sentences, would it have any reason to mention us — and would what it says be accurate?”

What it costs

Our AEO and AI-search visibility programmes start at ₱42,000/mo, usually alongside a technical SEO audit at ₱55,000 in month one. It compounds slowly and is best judged over two quarters, not two weeks — the same as traditional SEO, for the same reason. See SEO and discoverability for the full list.

FAQ

AI search questions

  • Answer Engine Optimisation is the practice of making a business quotable by AI assistants such as ChatGPT, Claude, Perplexity and Google AI Overviews, rather than only ranking in traditional search results. It relies on unambiguous factual content, structured data, AI crawler access, and a machine-readable summary such as an llms.txt file.

  • There is no meaningful difference — they are three names for the same discipline. AEO stands for answer engine optimisation and is the oldest and most widely used term. GEO stands for generative engine optimisation and is the term used in the academic literature. GSO stands for generative search optimisation and is the newest of the three. All three describe the work of becoming the source an AI assistant quotes: unambiguous factual content, structured data, crawler access for AI bots, and a machine-readable summary of the business. Anyone drawing a hard distinction between them is selling one of the labels.

  • No. AI answer selection is opaque and changes without notice, so guaranteed AI citations cannot be honestly promised. What can be done is removing every technical reason for exclusion — crawler access, structured data, quotable factual content, an llms.txt file — and measuring prompt-level visibility monthly.

  • An llms.txt file is a plain-text summary placed at a website's root that describes the organisation, its pricing and what each page covers, in a form AI systems can parse easily. It is an emerging convention that lets a business author the summary it would want quoted rather than leaving it to inference.

Want to be the source an assistant quotes?

A scoped plan, a fixed quote and a straight answer on what you don't need. We reply within one working day.