GEO Guide: Get Cited in AI Overviews, ChatGPT & Perplexity
Search is no longer only ten blue links. Google AI Overviews, ChatGPT search and Perplexity summarise the web and cite a few sources. GEO is how you become one of them.
Generative engine optimisation (GEO) is the practice of making your content easy for AI-powered search engines — Google AI Overviews and AI Mode, ChatGPT search, Perplexity, Microsoft Copilot — to understand, trust and cite in their generated answers. It builds on classic SEO: AI systems still retrieve pages through search indexes, so crawlable, well-ranking, clearly structured pages with direct answers are the ones that get quoted.
Key takeaways: rank first — AI answers mostly cite pages that already rank · put a direct 2–3 sentence answer at the top · use clear headings, lists and tables · add structured data · allow AI search crawlers · show real expertise and up-to-date facts · track AI referrals in analytics.
How AI search chooses sources
Most AI answer engines work in two steps. First they retrieve candidate pages for the query from a search index (Google's own, Bing's, or their own crawler). Then a language model reads those pages and writes a summary, citing the passages it used. To be cited you must be retrievable (indexed and ranking reasonably well) and extractable (the answer is easy to find and quote).
1. Write answer-first
Start each page and each section with a clear, self-contained answer — the "inverted pyramid". If the question is "What is a canonical URL?", the first sentence should define it in plain words. AI models prefer passages that answer completely without needing the rest of the page. Every tool page on URLMode follows this pattern: a short definition box directly under the tool, followed by the detailed guide.
2. Structure content for extraction
- Descriptive headings phrased like real questions ("How long should a meta description be?").
- Lists and tables for steps, comparisons and specifications — models lift them almost verbatim.
- Specific facts: numbers, limits, dates and definitions ("Sitemaps can contain up to 50,000 URLs").
- One topic per section, so a single passage can stand alone.
3. Add structured data
Schema.org markup in JSON-LD tells machines what your content is: Article, FAQPage, HowTo, Product, SoftwareApplication, Organization. Google limits some rich results, but structured data still helps every machine reader interpret your page. Create FAQ markup with the FAQ Schema Generator and validate JSON with the JSON Formatter.
4. Let the right crawlers in
Check your robots.txt. AI search features rely on crawlers such as OAI-SearchBot (ChatGPT search), PerplexityBot and Googlebot itself. You can block training-only crawlers (GPTBot, Google-Extended, CCBot) and still allow search crawlers. Test any rule with the Robots.txt Checker. Also make sure content is in the HTML — content that only appears after heavy JavaScript is harder for some AI crawlers to read.
5. Consider llms.txt
llms.txt is a proposed plain-text file at your site root that lists your most important pages with one-line summaries, written for language models. It is not an official standard and Google does not use it, but it costs little and can help other AI tools navigate your site. You can see URLMode's at /llms.txt.
6. Demonstrate experience and trust
- Original data, screenshots, examples and first-hand testing — things a model cannot invent.
- Clear authorship and an About page explaining who you are.
- Up-to-date information with visible "updated" dates.
- Mentions and links from reputable sites — AI systems lean on the same authority signals as search.
7. Measure AI visibility
- In GA4, look for referral sources such as
chatgpt.com,perplexity.ai,copilot.microsoft.comandgemini.google.com. - Google Search Console includes AI Overviews and AI Mode traffic in its Performance report.
- Regularly ask AI assistants your target questions and note which sources they cite.
- Track the pages that win citations and apply their structure to others.
GEO checklist
| Area | Check |
|---|---|
| Indexing | Page is indexable, canonical and in the sitemap |
| Answer | Direct 2–3 sentence answer at the top |
| Structure | Question headings, lists, tables, one idea per section |
| Markup | Article / FAQPage / Organization schema |
| Access | AI search crawlers allowed in robots.txt |
| Trust | Author, sources, dates, original insight |
Run a quick technical baseline with the SEO Audit — the foundation for both search and AI visibility.
Frequently asked questions
What is GEO in SEO?
GEO, or generative engine optimisation, is optimising content so AI search tools such as Google AI Overviews, ChatGPT and Perplexity understand it and cite it in their answers.
Is GEO different from SEO?
It builds on SEO. AI engines mostly retrieve pages through search indexes, so ranking well is still the foundation; GEO adds answer-first writing, clear structure and machine-readable markup.
Does llms.txt help with Google?
No. Google has said it does not use llms.txt. It may help some other AI tools navigate a site, and it costs little to add.
Should I block AI crawlers?
Block training-only crawlers if you do not want your content used for model training, but allow AI search crawlers if you want to be cited in AI answers.
How do I know if AI tools send me traffic?
Check GA4 referral sources for domains like chatgpt.com and perplexity.ai, and review Google Search Console, which includes AI Overview traffic.
Shorten links & create QR codes free
Clean fvj.io links, dynamic QR codes and click analytics — plus 90 free SEO tools.
