GEO — how to appear in ChatGPT, Perplexity & Claude.

GEO (Generative Engine Optimization) is the practice of optimising content so that AI models — ChatGPT, Perplexity, Claude and Gemini — cite your site as a source in their answers.

How does GEO differ from SEO?

Classic SEO fights for a click from Google results. GEO fights for a citation in an AI answer. The user doesn't click — they read the answer and trust the model. The winner is whoever the model picks as a source.

How do LLMs pick who to cite?

Based on analysis of 12,000 answers from ChatGPT, Perplexity, and Claude (FETCHER Research Q1 2026), 7 repeatable signals stand out:

1. Clear authorship — article has a specific author with name and bio
2. Quotable sentences — concrete facts, numbers, dates
3. Structured data — Schema.org (Article, FAQPage, HowTo) boosts citation chance 3×
4. Freshness — articles under 6 months old are preferred for time-sensitive queries
5. Expertise — LLMs recognize authors appearing across multiple reputable sources
6. Technical availability — site renders without JS (LLM crawlers often skip JS)
7. llms.txt — a root-level file telling LLMs how to treat your content

Which GEO tactics work in 2026?

1. Write in "question → answer" format

LLMs love Q&A structure. Use H2/H3 as questions, answer in the first paragraph, then expand. Boosts direct citation chance up to 5×.

2. Add llms.txt

Markdown file at root (`/llms.txt`) tells LLMs which content matters:

# FETCHER Solutions

> Tech agency from Poland specializing in AI automation, websites, Shopify stores.

## Key pages

- [Services](https://fetchersolutions.pl/en/#services)
- [Pricing](https://fetchersolutions.pl/en/#pricing)

## Expert articles

- [Automation 2026](https://fetchersolutions.pl/en/blog/business-process-automation-2026.html)
- [GEO Guide](https://fetchersolutions.pl/en/blog/geo-optimization-chatgpt-perplexity.html)

3. Schema.org — maximum types

• Article + author + datePublished on blogs
• FAQPage for Q&A
• HowTo for step-by-step
• LocalBusiness for local
• Organization site-wide
• Product + Review for e-commerce

4. Build author authority

Every article should have author name (not "Admin"), link to bio/LinkedIn, Schema `Person` with `sameAs`, consistency — same author writing in 5+ places online.

5. Facts, not opinions

LLMs don't cite opinions ("we think…"). They cite facts ("in 2026 the average Shopify Basic store costs $29/mo"). Convert subjective statements into measurable numbers.

6. Layered content structure

Build topic clusters: one long pillar + 5–10 shorter subtopics, all interlinked. LLMs map your expertise more easily.

7. Render without JavaScript

Perplexity and GPTBot crawlers often don't execute JS in 2026. Use SSR/SSG (Next.js, Astro) or static HTML. Client-rendered SPAs are invisible.

8. Get mentions, not just links

LLMs index brand mentions even without links (unlinked mentions). Getting mentioned in reputable sources (Forbes, TechCrunch, Wired) builds GEO authority even stronger than backlinks.

9. Optimize for "direct answer"

Write the first paragraph as a complete answer to the H1 question. LLMs often cite just the first 2–4 sentences.

10. Monitor GEO

Track citations in LLMs: Perplexity API shows sources, ChatGPT — prompt regularly for your brand/vertical, Google AI Overviews — monitor in Search Console, tools: Profound, Brightedge GEO, HubSpot LLM Tracker.

What should you avoid in GEO?

• Don't keyword-stuff — LLMs detect and penalize
• Don't mass-generate AI content — duplicate content, no authority
• Don't block LLM bots in robots.txt (unless you know what you're doing)
• Don't drop SEO — GEO adds to, doesn't replace. Classic SEO still drives 70% of traffic

GEO checklist 2026

1. ✅ Add llms.txt at domain root
2. ✅ Review and complete Schema.org on every page
3. ✅ Rewrite H2/H3 as questions
4. ✅ Add author to every article (with Person schema)
5. ✅ Verify site renders without JS (view-source)
6. ✅ Add FAQ with 5+ questions (with FAQPage schema)
7. ✅ Test your brand in ChatGPT / Perplexity — see what they know
8. ✅ Build 3 topic clusters
9. ✅ Get 2 mentions in reputable industry outlets
10. ✅ Monitor citations monthly

Takeaways

GEO won't replace SEO in 2026 — it'll coexist. But companies ignoring AI search will see 30–50% organic traffic drop in 2–3 years. GEO rollout is fast (mostly tidying what you have) and cheaper than a classic SEO campaign. So — start today.

Frequently asked questions

What is GEO?

GEO (Generative Engine Optimization) is optimising content to be cited in answers from AI models such as ChatGPT, Perplexity and Claude. Classic SEO fights for a click from Google results; GEO fights for a mention in the answer itself, because the user reads the model's response and trusts the sources it picked.

Will GEO replace SEO?

No. GEO adds to SEO rather than replacing it — classic SEO still drives most traffic, and the two will coexist. GEO rollout is fast, though, because it is mostly about tidying what you already have, and it is cheaper than a classic SEO campaign. Effects typically show within 2–6 months.

What is an llms.txt file?

llms.txt is a Markdown file placed at the root of your domain (/llms.txt) that tells language models which content on your site matters most — for example services, pricing and key expert articles. It is one of the signals that helps LLMs understand how to treat your content.

Which structured data helps with GEO?

Each page should carry the right Schema.org type: Article with author and datePublished on blogs, FAQPage for Q&A, HowTo for step-by-step guides, LocalBusiness for local firms, Organization site-wide, and Product plus Review for e-commerce. Add Person schema with sameAs links for your authors.

How can I check whether AI engines cite my site?

Track citations regularly: Perplexity shows its sources, ChatGPT can be prompted periodically about your brand and vertical, and Google AI Overviews queries can be monitored in Search Console. Tools such as Profound, Brightedge GEO and HubSpot LLM Tracker also help. Review your citations monthly.