In a recent LinkedIn post, Ann Smarty has shed light on a previously underutilized aspect of Google Search Console, revealing how businesses can gain crucial first-hand insights into how users are prompting AI platforms like ChatGPT, Claude, and Gemini. For years, understanding the average customer’s interaction with these AI tools has been a significant challenge, but Smarty’s discovery offers a direct path to this valuable data.
Smarty’s analysis points to a specific area within Search Console where these AI-driven queries are being recorded. “It was recently noticed and confirmed that AI Mode records those prompts as queries in the ‘Performance’ section of the Search Console,” she states, highlighting a significant shift in how search behavior can be monitored.
“This was noticed by ‘yes’ queries, which were confirmed to be follow-up questions in AI Mode. I see those as well…”
Gaining Actionable Intelligence from AI Queries
The implications of this discovery are substantial for content creators and SEO professionals. By examining these recorded AI prompts, businesses can understand which of their existing assets are being leveraged by Large Language Models (LLMs). As Ann Smarty explains, this information is immediately actionable.
“Looking at the clients’ reports, this idea already looks useful, as I can click to the ‘Pages’ tab for each of those queries to see which assets LLMs are pulling from the client’s website. Lots of additional opportunities there (update, optimize, create similar ones, etc.)”
Smarty emphasizes that this allows for strategic updates and optimizations of content that is already proving relevant to AI-driven searches. This proactive approach can help businesses stay ahead by refining content that is resonating with AI models, thereby potentially improving their visibility in AI-generated responses.
Advanced Techniques for Uncovering AI Prompts
Beyond the basic identification of AI queries, Ann Smarty also suggests advanced methods for uncovering more nuanced interactions. She proposes using a specific regular expression (regex) to identify longer, more conversational prompts.
Utilizing Regex for Deeper Insights
The regex provided by Smarty is designed to capture prompts that are at least ten words long, offering a more detailed view of user intent when interacting with AI. “Another idea is to pull longer, conversational queries using the following regex: ([^” “]*s){10,}?” she shares. This technique can be particularly useful for understanding complex user queries that might not be immediately apparent through standard Search Console metrics.
Important Considerations for Data Access
However, Ann Smarty also includes a crucial caveat regarding data visibility, especially for high-traffic websites. She points out that for properties with significant traffic, Search Console will only display queries that have generated at least some clicks. This means that prompts which do not result in any clicks might be missed.
“Note: This must be common knowledge, but since I was confused about that as well, here’s your reminder… For high-traffic properties, you will only be able to see queries that send at least some clicks, only. Since AI Mode might be sending none, you will miss those unless you access the SC data through the API.”
This limitation underscores the importance of utilizing the Search Console API for a more comprehensive data set, particularly when seeking to capture every interaction, even those that don’t immediately drive traffic. Smarty’s insights provide a valuable, practical guide for marketers and SEO professionals navigating the evolving landscape of AI-driven search and content interaction.
📝 About This Content
This article is based on insights shared by Ann Smarty on LinkedIn.
📅 Originally posted on August 8, 2026 | View original post on LinkedIn →