-
Claude’s citation behavior is genuinely different from ChatGPT and Gemini, it tends to be more conservative and pulls from a narrower set of sources. The most common reason I see for a site that ranks well everywhere else getting ignored by Claude is that Claude is more reliant on structured, extractable content from domains it considers authoritative, and it has a stronger preference for content that includes explicit comparison frameworks (tables, pros/cons, verdicts) rather than narrative roundups.
If your site ranks #1 on Perplexity and ChatGPT, you likely have good content and schema. But Claude may be drawing from a different training data snapshot or weighting citation sources differently, it is known to favor Wikipedia, established media, and domains with very clean structured data over newer or niche sites, even if those sites have strong topical authority elsewhere.
One thing worth checking: whether Claude is citing a specific competitor’s domain that has a Wikipedia entry or a G2 profile with high review volume. Those signals carry disproportionate weight in Claude’s responses compared to ChatGPT. If that tiny site is owned by the same company, it might have a Wikipedia page or a long-standing domain history that Claude’s model treats as more authoritative.
The fix that has worked for people I have talked to is not more content, it is getting your domain cited on the sources Claude actually pulls from. A Wikipedia entry if you qualify, or getting mentioned in a Wirecutter/CNET-style comparison that Claude indexes, tends to move the needle more than optimizing your own pages.
