Optimising Content for ChatGPT and Perplexity: the Practical GEO Guide
7 concrete techniques to appear in ChatGPT, Perplexity and Google AI Overview responses: structure, sourced data, entities, marked-up FAQ.
Optimising Content for ChatGPT and Perplexity: the Practical GEO Guide
Ranking well on Google is no longer enough. A growing share of your potential clients now gets their answers directly from ChatGPT, Perplexity, or Google AI Overview — without ever clicking on a site. If your content isn't designed to be cited by these AI engines, you're losing visibility you don't even see leaving.
This practical guide gives you concrete techniques for optimising your existing content — and creating new content — to appear in responses generated by LLMs.
Understanding How LLMs Choose Their Sources
Before optimising, you need to understand the selection mechanism.
When Perplexity answers a question, here's what happens behind the scenes:
- The user's query is analysed
- Perplexity performs several Google/Bing searches in parallel
- It selects the most relevant pages from the results
- It reads and extracts the most informative passages
- It synthesises these passages into a coherent response with citations
ChatGPT Search works similarly. Google AI Overview uses both its own index and its own AI models.
What this means for you: if you're not in the top results for a given query, you have virtually no chance of being selected as a source by an LLM. Classic SEO remains the indispensable prerequisite.
But among ranking sites, the AI performs an additional selection based on the readability, factuality, and structure of the content.
The 7 GEO Optimisation Techniques
Technique 1 — Answer Directly at the Start of Each Section
LLMs look for direct answers to questions. The ideal structure to get cited:
[H2] What is GEO?
[First paragraph] GEO (Generative Engine Optimisation) is the set
of practices that allow content to appear in AI engine responses
like ChatGPT or Perplexity.
The direct answer must come immediately after the section heading, within the first 2-3 sentences. The AI can then extract it cleanly.
Technique 2 — Structure with Lists and Tables
Bullet-point lists and tables are extraordinarily well absorbed by LLMs. They enable synthesis and answer "what are the X?" or "what's the difference between A and B?" questions.
A GEO vs SEO comparison table will be far more citable than a 200-word explanatory paragraph. A list of 5 best practices will be extracted in full rather than a continuous block of text.
Technique 3 — Include Sourced Statistics
LLMs trust sourced figures. "According to a Brightedge study (2024), 68% of online traffic begins with a search" — this type of precise, named-source data has a much higher chance of being included in an AI response than "most people start browsing with Google."
Recommended practice: for each important claim, find a recent study supporting it and cite it explicitly in the format (source: Name, Year).
Technique 4 — Create Structured FAQ Markup
FAQ sections have a dual advantage: they target Google's featured snippets (improving your SEO), and they're particularly well absorbed by LLMs that look for question-answer pairs.
Each question should be a real question your users ask (check in Google Search Console or Semrush). Each answer should be concise — 2 to 4 sentences maximum — and direct.
Add the FAQPage schema in JSON-LD to reinforce the semantic signal.
Technique 5 — Build Solid Author Pages
LLMs value verifiable expertise. A well-built author page includes:
- Full name and photo
- Detailed bio with years of experience and areas of expertise
- Links to external publications and mentions
- LinkedIn profile and/or other professional networks
This author page should be linked from every article you publish. It builds your "entity" in LLMs' eyes.
Technique 6 — Sign Your Content with Named Entities
LLMs distinguish between "an anonymous article about SEO" and "an article signed by a specific person with documented expertise in that field." This difference is captured by the concept of entity.
Make sure that:
- Your name is correctly associated with your site in external mentions and backlinks
- You have a Wikipedia or Wikidata presence if possible
- Your LinkedIn, Twitter/X, and other network profiles are consistent with each other and your site
Technique 7 — Be Active on Platforms LLMs Crawl
Perplexity, among others, crawls:
- Reddit (discussions, subreddits of your niche)
- Quora (Q&A)
- YouTube (video transcripts are often indexed)
- Specialist forums
- Hacker News
A well-written response on a relevant Reddit thread, a Quora article on your expertise, transcripts of your YouTube videos — all of this can become a direct source for LLMs.
GEO Audit: How to Evaluate Your Current Visibility
Here's a simple process for evaluating where you stand:
Step 1 — Test your target queries. Take your top 10 SEO queries and ask them exactly to ChatGPT, Perplexity, and Google AI Overview (in incognito mode). Are you cited? Are your competitors cited?
Step 2 — Analyse the cited sources. When you're not cited, who is? Analyse the content of the selected pages. How do they differ from yours?
Step 3 — Gap analysis. Identify the missing patterns: fewer sourced statistics? Less clear structure? No FAQ? Non-existent author page?
Step 4 — Prioritisation. Start with pages closest to being cited — those ranking well in SEO but not being selected. Targeted optimisation may be enough.
GEO and SEO: the Complete Stack
GEO optimisation doesn't replace SEO — it complements it. SEO fundamentals (technical, quality content, backlinks) remain essential. GEO adds a layer:
- Structural clarity for machine reading
- Factuality for AI reliability
- Entities for personal credibility
- Multi-platform distribution for corpus presence
If you want to compare the two approaches in detail, read our article on SEO vs GEO and our guide on AI search.
Taking Action
You can start today with a quick audit: take your best blog post, ask the central question it addresses to Perplexity, and see if you're in the cited sources. If not, analyse why and apply the techniques above.
For expert guidance on your GEO strategy — AI visibility audit, optimisation plan, monthly tracking — contact me. And test your SEO intuitions in our simulator to better understand the trade-offs between SEO and GEO investments.
Article written by Alexandre De Sousa, SEO & GEO consultant, founder of iZZi. Last updated: May 2026.
SEO vs GEO: what's the difference, and do you have to choose?
Full comparison between classic SEO (Google) and GEO (AI engines). Summary table, concrete examples, and a guide to combining both approaches.
What is GEO? Understanding Generative Engine Optimisation
GEO (Generative Engine Optimisation) explained: why ChatGPT, Perplexity and Google AI Overview are changing the rules of search and how to adapt.
What is SEO? The clear definition nobody actually gives you
SEO explained plainly: the 3 ingredients of organic search (content, technical, authority) with a concrete example and links to dig deeper.
Ready to go further?
Get a free diagnostic of your real situation — I'll tell you in 20 minutes whether your site has a structural problem.