Generative engine optimization that gets you named in the answer.

Generative engine optimization is the work of getting your pages used and named as a source inside AI generated answers. Instead of competing for a link position, you are competing to be one of the few pages an engine pulls from when it writes an answer in ChatGPT, Perplexity, Claude, Gemini or Google AI Overviews.

Four things. It makes sure retrieval crawlers can reach your pages. It restructures content so an answer can be lifted out cleanly. It builds the entity footprint that makes an engine confident naming you. And it measures how often you are actually mentioned. The first and last are where most programmes quietly fail.

Answer engine optimization is the older sibling of GEO. It began as writing pages so a direct answer could be pulled into a featured snippet, a People Also Ask box or a voice reply. That same extractable structure now feeds AI answers, so AEO and GEO have largely converged.

LLM SEO is the same work named after the model rather than the surface. Someone saying LLM SEO usually means getting cited by ChatGPT, Claude, Gemini or Copilot. There is no separate methodology behind the label. Treat a vendor who claims otherwise as a marketing problem, not a technical one.

Partly, and being straight about that matters. Google states there are no additional requirements or special optimizations to appear in AI Overviews or AI Mode. So much of GEO is excellent ordinary SEO. What is genuinely new is per engine crawler access, extractable answer structure, and measuring mentions instead of positions.

GEO is about being a source inside a generated answer. AEO is about having a clean extractable answer on the page in the first place. AEO is a component of GEO, not a rival to it. Anyone selling them as two separate retainers is selling the same work twice.

Traditional SEO competes for a ranked position and is measured in clicks. GEO competes to be quoted inside an answer and is measured in mentions. The underlying index is largely shared, so the technical foundations overlap heavily, but the winning page shape and the reporting are different.

It depends what you need first. If your organic foundation is weak, start there, and our broader AI SEO services page covers that programme. If you already rank and are simply not being named in AI answers, this page is the right starting point. Most brands do both, in that order.

Not on its own, and we will not tell you otherwise. Google states there is no special schema.org structured data you need to add to appear in its AI features. We still ship correct schema because it is cheap and helps entity resolution. We just do not sell it as a citation lever.

It is harmless and takes minutes, so we ship one. It is not a lever. No major engine has published a commitment to reading it. If an agency puts llms.txt at the centre of its GEO pitch, that tells you how much measured work sits behind the rest of the pitch.

The ones that retrieve live web content and cite sources: ChatGPT search, Perplexity, Claude, Google AI Overviews and AI Mode, Bing chat answers, and Apple surfaces fed by Applebot. We cannot change what a model memorised in training. We can change what it retrieves and links at answer time.

Access is the first gate and OpenAI documents it. Sites opted out of the OAI-SearchBot crawler will not be shown in ChatGPT search answers, though they can still appear as navigational links. That makes robots.txt the cheapest and most damaging mistake in the whole discipline.

It depends which bot. Blocking a training crawler such as GPTBot or ClaudeBot is a rights decision with no documented effect on search citation. Blocking a retrieval crawler such as OAI-SearchBot, Claude-SearchBot or PerplexityBot removes you from that engine. Many sites block both by accident with one broad rule.

Google-Extended is a robots.txt control token, not a crawler. Google documents that it has no separate user agent, and that it does not affect inclusion in Google Search and is not a ranking signal. It governs training and grounding in Gemini apps and the Vertex AI API. Allowing it is a rights call.

By doing nothing, according to Microsoft. Bing published that no action is needed to remain in its chat answers, and that content tagged NOARCHIVE will not be included in them or linked from them. If someone added NOARCHIVE years ago for another reason, it is costing you answer inclusion now.

We do not assume they do, and neither should you. Google tells site owners to make sure important content is available in textual form, and the other retrieval crawlers publish nothing at all about rendering. So anything load bearing gets server rendered. If your key content only exists after hydration, treat it as invisible.

We start with an access and extraction audit: every retrieval crawler token checked by name, rendering checked with JavaScript off, and your top pages scored for whether an answer can be lifted out of them. Alongside that we run a baseline mention sweep so we know where you stand.

Mostly rewrite. A page that already ranks and already has links is a far better citation candidate than a new page with neither. We restructure it into extractable units, raise the density of named systems and standards, add a real FAQ block, and fix the schema. New pages come later.

Self contained factual sentences that survive being lifted out of context. Real list structure instead of long prose. Specific named systems and standards rather than generic claims. An honest last updated date, and a named author. Those are the traits that recur across the pages we measured as cited.

Crawler access fixes can register within about a day, since OpenAI documents that robots.txt changes take roughly twenty four hours to be picked up. Content and entity work is slower and usually shows in months rather than weeks. Anyone promising ChatGPT citations in thirty days is guessing.

We run the same buyer questions repeatedly across ChatGPT, Claude, Gemini and Perplexity and record how often you are mentioned, where in the answer, and which competitor domains own it instead. Repeating the prompt matters, because these answers vary run to run. One screenshot is not a measurement.

No, and a guarantee would be a red flag. Answers vary between runs on the same prompt, and no engine publishes its selection criteria. What we commit to is a measured baseline, a specific set of changes, and a repeat measurement on the same prompts so you can see whether it moved.

The visit count is the wrong metric. An answer that names you reaches a buyer at the moment they are building a shortlist, and it works even when nobody clicks. Judge it on mention share across your buying questions, and on how many sales conversations open with a recommendation.

A short call where we look at your robots.txt, your top pages, and the questions your buyers actually ask an assistant. You leave knowing whether you are blocked, whether your pages are extractable, and what a realistic programme looks like. You can also run our free AI visibility check first.