ChatGPT and Conversational Search: Impact on SEO

Where Conversational Answers Appear

Google’s guide to generative AI features says you don’t need special files: llms.txt and similar “special” markup aren’t required for its features. Its AI features documentation adds that there’s “no special schema.org structured data that you need to add.” How AI Overviews work in Search is covered in AI Overviews.

Crawlers: Search and Training Are Different

AI companies use separate crawlers for different jobs, and blocking one doesn’t block the others. OpenAI’s bot documentation describes them:

• OAI-SearchBot “is for search”: “Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links.”
• GPTBot crawls content “that may be used in training our generative AI foundation models.”
• ChatGPT-User visits a page when a user’s request needs it; OpenAI says it “is not used for crawling”.

Google has a similar split. Google-Extended is a robots.txt token, not a separate crawler, that controls whether content Google crawls may be used for training future Gemini models and for grounding in Gemini Apps and Vertex AI, per Google’s crawler documentation. Blocking it doesn’t remove a site from Google Search. Before changing robots.txt, decide separately whether you want to appear in AI search answers and whether you allow training; see robots.txt.

What Changes for Content

• Short factual questions can be answered in the response itself. A page that only restates a definition has less reason to be visited.
• Questions that need judgment, comparison or detail can still lead people to sources: the answer can cite pages, and people may click through to check, compare or act.
• Actions happen on websites: buying, booking, calling, signing up. Pages built for the action keep their role.
• First-hand information (your own data, tested results, real examples, clear prices and policies) gives an answer something to cite that generic pages don’t.

Writing that answers questions directly, in self-contained passages with precise terms, serves both search results and conversational answers; see answer engine optimization and helpful content. How citation works on each AI platform is covered in generative engine optimization.

Freshness

Answers that draw on a search index reflect updated pages once those pages are recrawled and reindexed; answers generated only from a model’s training data lag behind. Google’s AI features use its index, and ChatGPT search uses pages surfaced through OAI-SearchBot. Keep facts, prices and policies current, and make sure updated pages can be crawled.

Measuring Conversational Search Visibility

• Google’s AI features: clicks from AI Overviews and AI Mode are counted inside Search Console’s Web search type, and Google has added generative AI performance reports to Search Console; see the Google Search Console guide.
• Other assistants: GA4 has an AI Assistant channel; its channel help describes it as arrivals “from sources like ChatGPT, Gemini, Deepseek, Copilot, or Grok” and notes that it “excludes Google’s AI Overviews and AI Mode.” Setup is in GA4 for SEO.
• Citations: for your priority questions, ask the assistants yourself and record which sources they cite, with the date. Results vary by wording and over time, so track a fixed set of prompts.
• Brand demand: watch branded searches in Search Console, since people who meet your brand in an answer may search for it later.

Responding Without Overreacting

Traffic effects differ by query type and site, and published estimates vary; measure your own before changing strategy. Keep the fundamentals that every surface depends on: indexable pages, clear answers, accurate facts and a reason to visit. Longer-term changes are discussed in the future of search.

Write the robots.txt decision as separate lines for OAI-SearchBot and GPTBot, and check that a rule meant to stop training didn’t also stop search: a site that disallows OAI-SearchBot is left out of ChatGPT search answers, not only out of training.