AEO & GEO
Answer Engine Optimization
More and more people ask an assistant instead of a search engine, and the assistant answers from the handful of sources it could parse, trust and retrieve. Answer Engine Optimization (AEO) — also called Generative Engine Optimization (GEO) — is the work of becoming one of those sources.
This is an engineering problem before it is a marketing one. AI systems crawl, chunk, embed and rank your content, so what decides the outcome is the machine-readable layer underneath the design: a connected schema.org entity graph, sections that still make sense after being cut into chunks, factual claims stated where a retriever can actually find them, and clean access for the AI crawlers themselves.
What I do
- AI visibility audit: what assistants say about you today, and the technical reasons why
- schema.org entity graph — Organization, Person, Service, FAQ and breadcrumbs, linked by @id
- llms.txt and llms-full.txt, so assistants can read your site instead of scraping it
- Retrieval-ready content architecture: chunk-safe sections, answer-first phrasing, real anchors
- Crawler and indexing plumbing: a correct robots.txt, IndexNow submission, honest sitemap dates
- Measurement: tracking citations and assistant-referred traffic over time
Why me
I build the systems that do the answering. RAG pipelines, embeddings, semantic search and LLM evaluation are my day job as a Senior AI Solutions Engineer, so I optimize for how retrieval actually behaves rather than for a checklist. This site is the reference implementation: llms.txt and llms-full.txt, a linked JSON-LD graph, IndexNow submission and sitemap dates that only move when content really changes.
Frequently Asked Questions
How is AEO different from SEO?
They overlap, and neither replaces the other. Classic SEO decides whether you rank; AEO decides whether an assistant can parse, trust and quote you. Clean HTML, structured data and fast pages serve both — AEO adds machine-readable summaries, entity linking, and content written to survive being cut into chunks.
Can you guarantee that ChatGPT will recommend me?
No, and anyone who guarantees it is selling something. These systems have no ranking dial to turn. What I can do is remove every technical reason an assistant would skip or misread you, and make the claims you want quoted explicit, verifiable and easy to retrieve.
How long before it shows up in answers?
Technical changes are picked up in days to weeks: IndexNow notifies Bing (which feeds Copilot) within minutes, while other crawlers re-read on their own schedule. Visible change in assistant answers usually takes weeks and also depends on how much corroborating information about you exists elsewhere on the web.
What do I actually get?
A working implementation in your codebase, not a PDF: the schema graph, the llms.txt files, crawler configuration and content changes — plus a short written rationale for each decision, so your team can maintain it after I'm gone.