Claude’s visibility tied to Brave Search rankings, data suggests
▼ Summary
– Claude uses Brave Search’s top 10 results directly without re-ranking, and it searches the web only 36.6% of the time compared to ChatGPT’s 90%.
– Claude is most likely to search for prompts involving recency (81%), rankings (67%), location (55%), or comparisons (51%), while prompts like “how does” or “what is” rarely trigger search.
– Claude’s citations overlap with Google rankings 64% of the time but only 8% with ChatGPT, suggesting Google SEO efforts may transfer to Claude more effectively.
– Claude’s query fan-outs are nearly deterministic, producing the same fan-out 65% of the time, often including years, giving an advantage to page titles with current-year signals.
– Because Claude’s search behavior is consistent and closely tied to observable Brave rankings, it may be one of the most optimizable AI answer engines for SEO.
New findings suggest that Claude’s visibility in AI-powered search results is more tightly linked to Brave Search rankings than other competing AI answer engines. Jonathan Clark, managing partner at Moving Traffic Media, shared this insight on LinkedIn after a session from Zero Click by Profound.
During the session, the standout takeaway was that Claude does not re-rank search results. Instead, it pulls directly from Brave’s top 10 results when generating answers. This behavior sets Claude apart from other AI tools like ChatGPT.
Claude also searches the web far less frequently. According to Clark, Claude used web search in only 36.6% of prompts, compared to roughly 90% for ChatGPT. The AI was most likely to search when prompts indicated freshness, rankings, location, or comparison intent. For example, recency-driven queries like “best XYZ” triggered a search 81% of the time, while ranking-focused prompts did so 67% of the time. Location-based prompts triggered search 55% of the time, and comparison queries such as “X vs. Y” triggered it 51% of the time.
The weight of Brave rankings is significant. Clark noted that Claude’s citations overlapped with ChatGPT’s in only 8% of cases when responding to the same prompts. However, Claude’s results showed a much higher overlap with Google rankings at 64%. This suggests that Google SEO efforts may transfer more effectively to Claude than strategies aimed specifically at improving visibility in ChatGPT.
This finding also elevates the importance of tracking Brave rankings. Clark emphasized that Claude relies on Brave, and ranking well in Brave gives marketers “something we can monitor and correlate to data.”
Certain prompts are less likely to trigger a web search. Queries starting with “how does,” “what is,” or “steps to” often keep Claude from searching, meaning it cannot cite web pages. Conversely, terms like “best,” “top,” “near me,” and comparison-style queries drove the highest search rates.
Clark also observed two patterns that could make Claude easier to test. First, Claude’s query fan-outs were nearly deterministic, producing the same fan-out 65% of the time across different users. Second, these fan-outs frequently included years. This means page titles with current-year signals may gain an edge in Claude-triggered searches, especially for ranking and freshness-driven prompts.
Why this matters. Claude’s visibility appears to depend more on ranking in the search results it uses. Clark’s conclusion was that Claude may be one of the most optimizable AI answer engines today because its search behavior is more consistent and more closely tied to observable search rankings.
(Source: Search Engine Land)
