AI search
A website usually fails to appear in ChatGPT answers for one of four reasons: its crawler is blocked in robots.txt, the page depends on JavaScript that the crawler does not execute, the content does not state the answer and the business facts clearly, or the site has too little independent evidence for a model to trust it. Fixing crawl access and publishing clear, structured, factual content addresses the first three; reputation and citations address the fourth.
Open your robots.txt and search for GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, and Google-Extended. A wildcard `Disallow: /` blocks them too. Some sites block AI crawlers through a CDN or firewall rule rather than robots.txt, which the file will not show.
Use the free AI visibility check to see exactly which crawlers are allowed and which rule decided it. Blocking training crawlers can be a legitimate choice; blocking search and citation crawlers usually removes you from the answer.
Answer engines work from text they can retrieve. If the main content only appears after client-side JavaScript, a canvas, or a login, a crawler may see an empty shell. Server-rendered HTML, static pages, and clean semantic headings all improve the odds that the page is understood.
The same applies to facts. A page that says 'we are passionate about excellence' gives a model nothing to quote. A page that says what the company does, who it serves, where it operates, what it costs, and how to start is citable.
Models assemble answers from multiple sources and tend to favour entities they can verify. A company page is one signal; consistent profiles, reviews, directories, published articles, and mentions on other sites are others. This is the part that takes time and cannot be fixed by markup alone.
Start with the entity basics: the same company name, description, address, and links everywhere. Then publish genuinely useful material that other people cite. There is no shortcut that reliably substitutes for being a real, consistent presence.
See which AI crawlers are allowed, whether llms.txt and structured data exist, and the highest-value fixes.
Last reviewed 22 October 2026. If this answer is out of date, tell us and we will correct it.
Generative engine optimization (GEO) is the practice of making a website's facts easy for AI answer engines to retrieve, understand, and cite. It combines technical crawl access, structured data, clear answer-first writing, and consistent entity information. Unlike traditional SEO, which competes for a link position, GEO competes to be included in a generated answer.
To get found by AI search, allow AI crawlers, publish clear answer-first content with explicit facts, add structured data, keep your company details consistent everywhere, and earn mentions on other trustworthy sites. Technical access gets you read; clarity gets you quoted; corroboration makes you worth citing.
A technical website audit reviews how a site is built and delivered: crawlability, indexing, response performance, Core Web Vitals, accessibility markup, HTTPS and security headers, structured data, broken links, and the reliability of forms and integrations. A serious audit separates what was measured from what was inferred, and states what it did not test.