How to get cited by ChatGPT: five moves, in order of leverage

1. Let the crawlers in

GPTBot, ClaudeBot, PerplexityBot and Google-Extended each obey robots.txt. Plenty of sites blocked them in 2023 out of caution and forgot. If the crawlers cannot read you, nothing else on this page matters. Check your robots.txt today, it takes ninety seconds.

2. Own specific answers

Engines do not cite "thought leadership". They cite the page that cleanly answers the exact question asked. One page, one question, answered completely, beats one sprawling guide gesturing at twelve. Build pages for the questions your buyers actually ask an engine, including prices, which almost nobody publishes and engines love to quote.

3. Structure for quotation

Put the complete answer in the first paragraph, then elaborate. Engines lift openings. If your answer starts after 400 words of scene-setting, the citation goes to someone less shy. Definitions, numbers and comparisons in parseable structures, tables over prose walls, schema on everything that has a type.

4. Exist where the models look

Engines cross-reference. A brand that appears only on its own domain is a rumor; one confirmed by directories, review platforms, press and industry lists is a fact. This is why earned links and mentions now do double duty: authority for Google, corroboration for AI engines.

5. Monitor, because engines drift

Models update, sources shift, and last quarter's accurate description becomes this quarter's wrong price or dead product name. Run the same buyer prompts monthly across engines, track cited versus mentioned versus invisible, and fix what drifted. That loop is exactly what AI Citation Monitoring runs for you.

Not sure where you stand today? Start with the baseline: the AEO Visibility Audit scores your presence across every major engine against competitors. And if you are new to the discipline, read what AEO is first. Perplexity is different enough to deserve its own page: how to get cited by Perplexity covers its two crawlers, its conflicting robots.txt guidance, and the one result you can verify in your own analytics.

A 30-day starter plan

Week one: check robots.txt against the five AI crawlers and fix access, then ask the four big engines your five most valuable buyer questions and record who gets named. Week two: pick the one question you deserve to own and rebuild that page answer-first, definition in the opening sentence, structure a machine can lift. Week three: fix your entity, same name, same description, same facts across your site, LinkedIn and the two directories that matter in your niche. Week four: re-run the five questions and record the delta. That loop, repeated monthly, is the whole practice.

Why deserving the citation still decides it

Every mechanical move above raises the probability of citation; none manufactures it. Engines are ranking trustworthiness at answer-time, and they lean on the same underlying signals search always has: who links to you, who mentions you, whether your claims check out. The mechanics get you considered. Being genuinely the best answer to a specific question is what gets you chosen, which is inconvenient and also the entire opportunity.

Sources

OpenAI documents more of this than people expect. See the overview of OpenAI crawlers and the publishers and developers FAQ. For the equivalent on Google's side, read Google's generative AI optimization guide.

Common questions