Claude does not draw evenly from the web. It skews toward well-established, text-dense sources that were prominent at training time. Claude answers from model weights unless web search is enabled for the conversation. For your own category rather than the published average, Cituna checks daily which brands Claude recommends and how often you are among them.
How Claude actually decides
In Claude’s default mode there is no live retrieval at all, so what Claude says about your brand reflects what was in its training data, which is why the fix for Claude is slower and more structural than on the live-retrieval engines.
In practice that means the surfaces worth auditing first are documentation sites, long-form editorial, Wikipedia, technical references.
The method, in order
Start from the skew, not from your site. Claude’s citations concentrate in documentation sites, long-form editorial, Wikipedia, technical references.
Separate the sources it cites from the sources it trusts silently. This engine frequently asserts without linking, so the citation list understates what shaped the answer.
Read the citations prompt by prompt, not as one total. A source that appears on your buying questions matters more than one that appears on definitions, and a single count hides the difference.
Check your category specifically. Citation skew is a population-level finding; the sources for legal advice and for developer tooling are not the same, and only your own prompt set tells you which applies.
Audit whether you appear on the top sources at all, then rank the gaps by how hard each is to close: a subreddit comment and a Wikipedia citation are not the same amount of work.
Re-measure after each placement lands. Claude’s volatility is low in default mode; high once web search is switched on.
The honest caveat
A brand founded or renamed recently can be absent from Claude entirely while doing well everywhere else, and no amount of on-page work fixes that quickly. It is the clearest example of why a single-engine tool misleads.
Cituna checks Claude daily alongside Google AI Overviews, ChatGPT, Gemini, Google AI Mode, Perplexity and Grok, on the $39 Starter, and pairs the result with Google Search Console so a change here can be tied to real clicks. Then it does the work: it generates the fix for each gap, its AutoSEO writes and publishes the articles that answer the questions you are missing, and Claude or any other AI agent can work through the results over MCP.
Drafted with AI assistance from our own research and Search Console data, and reviewed by Rahul A before publishing. Rules and prices change; check the linked official source before you act.