Skip to content
News Research

New data shows Claude pulls citations from Brave Search's top 10 results at an 86.7% overlap rate

· by Pondero Newsdesk

The short version

Analysis published in June 2026 found Claude's web citations track Brave Search's top 10 organic results at an 86.7% rate, far above ChatGPT's overlap with Bing. The finding puts Brave rank tracking at the center of any serious GEO strategy.

New data shows Claude pulls citations from Brave Search's top 10 results at an 86.7% overlap rate

Analysis shared in June 2026 found that Claude's web citations overlap with Brave Search's top 10 organic results at a rate of 86.7%, research first confirmed by Profound in 2025 and amplified through a Zero Click by Profound session reported by Search Engine Land on June 12. The finding reshapes how GEO practitioners should prioritize search engine optimization work.

What the data shows

Jonathan Clark, managing partner at Moving Traffic Media, presented findings at a Zero Click by Profound session reported by Search Engine Land that clarify how Claude selects sources when it searches the web. Claude does not re-rank Brave's results. It draws from Brave's top 10 directly, making Brave rank the single most actionable lever for earning a citation in Claude's responses.

Claude triggered web search in only 36.6% of prompts in Clark's analysis, compared with roughly 90% for ChatGPT. When it did search, recency-focused prompts such as "best XYZ" triggered retrieval 81% of the time. Ranking-focused queries triggered search 67% of the time, location queries 55%, and comparison queries 51%.

Claude's citation choices also diverged sharply from ChatGPT's. The two models cited the same sources in just 8% of cases when answering identical prompts. Claude's results showed a 64% overlap with Google's top rankings, suggesting that strong Google SEO does carry over to Claude to a degree. But the more direct mechanism runs through Brave Search, which Profound's earlier research put at 86.7% citation overlap.

Why it matters

For content teams focused on AI answer visibility, the practical implication is direct. Any content outside Brave's top 10 for a given query is effectively invisible to Claude on that query, regardless of Google domain authority or ChatGPT optimization work. Clark's session framing was that Claude may be one of the most optimizable AI answer engines precisely because its retrieval behavior is consistent and tied to observable rankings.

Claude also cited fewer sources per answer than ChatGPT, typically choosing two to four per response. Clark noted that Claude's query fan-outs were nearly deterministic: the same reformulated queries appeared across 65% of users. Page titles that include the current year showed an advantage in Claude-triggered searches for ranking and freshness-driven prompts. Prompt types such as "how does," "what is," and "steps to" were least likely to trigger a web search at all, limiting citation opportunities for purely definitional content.

The 86.7% overlap figure means Brave rank tracking is now a core GEO diagnostic. Teams currently relying only on Google Search Console data are missing the most direct signal available.

Context

Brave Search operates an independent index separate from Google and Bing. Anthropic has not disclosed a formal data-sharing agreement with Brave, but Claude's consistent use of Brave as its web retrieval backbone has been documented across multiple third-party studies. ChatGPT routes its web searches through Bing, where overlap between cited results and Bing's top organic results runs considerably lower per Profound's comparative data.

The timing of the Profound session also coincides with a broader shift in how practitioners talk about GEO. Ranking well in Brave has historically been a secondary consideration for most SEO teams. The citation-overlap data pushes it into primary position for any brand whose audience asks Claude for recommendations.

What to watch next

Three developments are worth tracking. First, whether Brave announces any formal partnership or data-sharing arrangement with Anthropic that would make the retrieval relationship explicit. Second, whether GEO tools add Brave-specific rank tracking as a standard feature alongside Google and Bing monitoring. Third, whether a parallel overlap analysis for Google AI Overviews versus Google organic results produces comparable numbers, which would clarify how much of the Claude-Brave overlap is unique versus a general pattern in AI answer engines.

Sources