Claude’s visibility may depend heavily on Liver Search levels, new data suggests

Claude may be more directly integrated with Brave Search’s ranking than other AI response engines, according to information shared by Jonathan Clark on LinkedIn from the Zero Click by Profound session.

Clark, managing partner at Moving Traffic Media, said the key to the session is that Claude “doesn’t repost search results” and instead appears to be using the top 10 Brave results directly in his answers.

Claude searches slowly. Claude used web search for 36.6% of commands, compared to 90% for ChatGPT, according to Clark.

• Claude would likely search when the information indicated novelty, quality, location, or comparative intent. Recency-oriented information such as “best XYZ” triggered searches 81% of the time, while ranking-oriented information resulted in searches 67% of the time.
• Location-oriented information triggered searches 55% of the time, while comparative information such as “X vs. Y” triggered searches 51% of the time.

The bold standard carried weight. Claude’s quotes went over with ChatGPT’s only 8% of the time when he was responding to the same information, according to Clark.

• Claude’s results had the highest overlap with Google’s ranking, at 64%. This suggests that Google’s SEO efforts may flow more easily to Claude than strategies focused specifically on improving visibility on ChatGPT, according to Clark.
• The findings also raise the importance of following the Brave standard. Clark said Claude uses Brave, and that placement in Brave gives us “something we can monitor and correlate with data.”

Some commands remain in the memory. Commands such as “how to,” “what,” and “steps to go” are less likely to trigger Claude to search the web. If Claude doesn’t search, he can’t cite web pages. Claude searched mostly for information containing words like “best,” “top,” “near me,” and comparative-style questions, according to Clark.

The years were often seen. Clark also noted two patterns that would make Claude easy to examine:

• Claude’s question for followers is almost decisive, generating the same followers 65% of the time for all users.
• Fan outings often spanned years.
  • That means that the headlines of the current year’s signaled pages may be useful in Claude-triggered searches, especially for high-quality information and new leads.

Why do we care. Claude’s visibility seems to depend heavily on the delivery of search results that Claude uses. Clark’s take was that Claude may be one of the most useful AI response engines today because its search behavior seems to be very similar and closely related to the level of visual search.