How Claude finds and cites brands
Checked against Claude documentation on November 4, 2026
Ocita doesn't track Claude. It tracks ChatGPT, Gemini, Google AI Mode and Google AI Overviews. This guide explains how Claude behaves.
Claude can answer from what it learned in training, or it can search the web. This guide explains how that works and what you can control. Ocita doesn’t track Claude yet, so nothing here describes a feature of Ocita.
When Claude searches
Anthropic’s API docs say Claude decides when to search based on the prompt. It searches when a request depends on information that’s current or changing, including details about specific organizations or products that might have changed, and when someone explicitly asks it to look something up. For stable knowledge, it answers directly.
Citations are always on for web search in the API. Each one carries the source URL, title and a short excerpt of the cited text.
The three Anthropic crawlers
Anthropic describes three bots, and says they honor robots.txt.
- ClaudeBot collects web content that could help train its models. Restricting it signals that your future material should be left out.
- Claude-User retrieves pages when someone asks Claude a question. Disabling it may reduce your visibility in user-directed search.
- Claude-SearchBot crawls the web to improve search result quality. Disabling it may reduce your visibility and accuracy in search results.
As with other engines, these are separate choices. You can restrict ClaudeBot for training and still allow the two that support search. Rules need setting on each subdomain, and Anthropic says blocking its IP addresses may not work reliably, because that stops it reading your robots.txt.
Search partners and noindex
Claude’s web search also relies on search partners. Anthropic says the noindex robots meta tag tells those partners not to index your content, and that content then won’t appear in Claude outputs that use web search. If a page matters to you, make sure it isn’t marked noindex by accident.
What to check first
- Read your robots.txt for rules naming ClaudeBot, Claude-User or Claude-SearchBot, and for any
User-agent: *rule that blocks them. - Check the same rules on each subdomain.
- Look for accidental
noindextags on pages you want found.
What Ocita does today
Ocita tracks ChatGPT, Gemini, Google AI Mode and Google AI Overviews. It doesn’t track Claude, so it can’t show you how Claude talks about your brand. The free AI crawler checker does read your robots.txt for ClaudeBot. Until Claude is tracked, ask it your category’s questions yourself a few times, and keep in mind that one answer proves very little.