The reference for making a website legible to LLMs, citable by AI answer engines and correctly interpreted in generative search. SEO, GEO and AEO in a single framework.
Editorial hub · LOOP method
Optimise your site for AI. With method.
Covers the engines ChatGPT Search Perplexity Google AI Overviews Gemini Claude
Choose a starting point
Four paths for your goals.
The same corpus supports different tasks. Start with the decision you need to make, then refine by engine, audience, format or evidence level.
Search the corpus- 01UnderstandDefinitions, differences between SEO, GEO and AEO, limits and terminology.Map the field →
- 02ImplementStructure, technical controls, crawlers, structured data and the LOOP method.Use the method →
- 03MeasureProtocols, prompt panels, observations and downloadable data.Open the observatory →
- 04VerifyEvidence levels, sources, corrections and the limits of each claim.Review the evidence →
The promise
A framework instead ofa pileof tips.
Generative search (ChatGPT Search, Google AI Overviews, Perplexity, Gemini, Claude) captures a growing share of acquisition. The promises flourish: "rank first on ChatGPT", "hack the AIs".
This site takes the opposite stance. Optimising for LLMs is a discipline of editorial engineering, not a sleight of hand. We publish here a structured methodology (LOOP), an operational 40-point audit, a clear reading of the concepts and applications by sector.
LOOP is our proprietary framework, documented publicly. The resources are dated, sourced and kept up to date.
The hub · 7 pillars
Seven pillars,one single method.
A topic cluster built to cover the whole subject: conceptual foundations, operational method, content engineering, technical architecture, framework, deliverable and proof by sector.
01
Foundations
What is LLM Optimisation?
Definition, scope, business stakes and real limits. What GEO, AEO and LLM SEO cover, and what they do not.
Read the page02
Method
How to optimise a site for AI
A six-step method to make a site visible and citable by AI answer engines, with no empty promises or magic recipes.
Read the page03
Content
Content structure for LLMs
Content design, chunking, self-containment, citation-friendly content. Editorial engineering applied to retrieval, without sacrificing human reading.
Read the page04
Technical
Technical optimisation
Crawl, rendering, schema.org, llms.txt, AI bot management (GPTBot, PerplexityBot, ClaudeBot, Google-Extended). The indispensable foundations.
Read the page05
Framework
The LOOP methodology
The four-dimension framework: Legibility, Ontology, Operations, Performance. A structured method to steer AI visibility as a loop.
Read the page06
Deliverable
Audit and checklist
Forty operational audit points to measure your site's LLM legibility, prioritise corrective actions and track progress.
Read the page07
Proof
Use cases
Concrete applications in B2B SaaS, e-commerce, media and professional services. Differentiated priorities by profile and sector.
Read the pageLOOP framework · Four dimensions
Four dimensions.One steering. A loop that turns.
L
Dimension 01
Legibility
Semantic structure, clean HTML, self-contained passages, accessibility. What the machine can extract without friction.
O
Dimension 02
Ontology
Entities, disambiguation, topical authority. What your brand means to an LLM and how it sits in its graph.
O
Dimension 03
Operations
Crawl, rendering, schema.org, llms.txt, AI bots. The technical foundations of a site indexable by AI engines.
P
Dimension 04
Performance
Measurement, monitoring, iteration. Realistic KPIs, citation tools, regular reporting. The loop closes here, and starts again.
Pre-flight · Checklist
Audityour site.
Run your site through the LOOP audit checklist. Forty points organised by dimension, clear deliverables, built-in prioritisation. Measure before you optimise.
Open the checklist →01
Semantic HTML5 architecture main, article, section, nav
Legibility
02
Correct heading hierarchy single H1, levels respected
Legibility
03
Self-contained passages up top TL;DR, definition lead
Legibility
11
Schema.org Organization, Article, FAQPage validated JSON-LD
Ontology
14
Entity disambiguation sameAs Wikipedia, Wikidata, LinkedIn
Ontology
22
llms.txt present and current domain root
Operations
25
AI bots explicitly allowed GPTBot, PerplexityBot, ClaudeBot, Google-Extended
Operations
38
AI citation tracking in place reference queries, frequency
Performance
and 32 more points in the full checklist. See the 40 points →