In this new era of productive AI technology, searchers have started replacing keywords with instructions. Short and long-tailed questions are being replaced by more conversational messages, which are often longer and more in-depth. These days, searchers expect complete answers rather than a list of results included in the papers.
Until we get a direct AI equivalent of Google Search Console or Bing Webmaster Tools, we can’t really see how our audience behaves on AI search platforms as they look for our content, products or products.
However, we can still look at proxies to simulate how this journey works. Here are several ways to use other data points as proxies to find the commands your audience uses. You can use your AI tracking tool to track how well these instructions are working.
People Are Asking Again
Hidden in plain sight, you can use the most popular SERP feature from keywords to notifications/questions. People Ask Again (PAA) was launched in 2014 and raises many questions related to your query. You can easily go from keyword to query list. If you click on any PAA result, the list expands and offers more terms.
Go question-by-question to find relevant PAAs, or you can use AlsoAsked to remove specific questions from the scale. PAAs are long questions that try to answer the next questions asked in the search journey, so they are a step closer to written instructions in the AI Search platform.
Userbots
Userbots like ChatGPT-User and Perplexity‑User are a powerful way to see how your pages are being used in AI Search. It doesn’t give you the instructions they use, but it will help you check which pages are being cited without trying to guess by following instructions that may or may not work at all.
These bots include URLs on your website when they are used to respond to the user. The process is called RAG (Retrieval-Augmented Generation).
Your customers are searching everywhere. Make sure it’s your product he appears.
The SEO toolkit you know, and the AI visibility data you need.
Start a Free Trial
Start with
To put it another way by Mike King, one of the most trusted sources in SEO & AI, RAG is a way in which a language model can be “grounded” in facts or learn from existing content to produce highly relevant outputs with little chance of fantasy.
Translation: Your page was used as feedback and, in some way or shape, your content helped the user. This can give you clues about what type of content your audience is using the most in these areas, even if the response hasn’t turned into a click.
Historically, log files have been difficult for SEOs to find, except that every website has them on their servers (another reason why SEOs access them!).
You can use a combination of pages with userbot visits, search for their keywords (as seen in Semrush or GSC), and see which PAAs Google has shown.
Longer questions on GSC/BWT
Despite not providing a breakdown of questions for AI Mode or AI Overview in GSC, smart SEOs find proxies that can be used to find questions that match the behavior we expect in these areas. One of them is Ziggy Shtrosberg, who came up with a great regex that you can copy and paste into your GSC.
His guidelines are:
- Sort by Search Appearance: Desktop
- Add a page filter with your root domain (eg, “https://www[dot]for example[dot]com/”) and add this big regex under the Query filter:
^(generate|create|write|make|build|design|develop|use|produce|help|assist|guide|show|teach|explain|tell|list|summarize|analyze|compare|give me|you have|you can|where|review|research|find|draft|compose|extract|process|convert|transform|plan|strategy|approach|method|framework|structure|overview|summary|breakdown|rundown|digest|perspectives|viewpoints|opinions|approaches|angles|pros and cons|advantages and disadvantages|benefits and drawbacks|assuming|suppose|imagine|consider|step by step|procedure|workflow|act as|adapt|prepare|advise|appraise|instruct|prompt|amend|change|advocate|aid|assess|criticise|modify|examine|your|assign|appoint|delegate|nominate|improve|expand|calculate|classify|rank|challenge|check|categorize|order|tag|scan|study|conduct|contradict|update|copy|paste|please|can you|could you|would you|help me|i need|i want|i'm looking for|im looking for|how do i|how can i|what's the|whats the|walk me through|break down|pretend you're|pretend youre|you are a|as a|from the perspective of|in the style of|format this as|write this in|make it|rewrite|i'm trying to|im trying to|i'm struggling with|im struggling with|i have a problem|i'm working on|im working on|what's better|whats better|which|pros and cons of|recommend|suggest|show me how|guide me through|what are the steps|how do i start|whats the process|take me through|outline the procedure|brainstorm|come up with|think of|invent|what if|lets explore|let's explore|help me think|i'm a beginner|im a beginner|as someone who|given that i|in my situation|for my project|i'm currently|im currently|my goal is|depending on|based on|taking into account|considering|given the constraints|with the limitation|improve this|make this better|optimize|refine|polish|enhance|revise|teach me|i want to learn|i don't understand|i dont understand|can you clarify|what does this mean|eli5|i'm confused about|im confused about|also|additionally|furthermore|by the way|who's|whos|find|more|next|also|another|thanks|thank you|please)( [^" "]*){9,}$
Take this strategy with a pinch of salt, as some of these questions may be done by LLM pursuers.
For example, I found an ad pattern that starts with “check,” which has a high number of impressions with zero clicks (not a minimum number of clicks, exactly zero clicks). If a long post has a high number of impressions and no clicks, be aware that it may not be people using these guides.
Confusion follow-up questions
One of the main AI Search platforms, Perplexity has a feature called “Related” where we display up to five tracking alerts. While initial information is still personal and may not be what others are promoting, relative tracking is still a good indicator of how fast people are—or at least what the platform expects people to do.
These answers are country specific, so do your research locally.
Considering that we don’t have a single keyword search volume metric and the instructions are more diverse than the keywords, it doesn’t make sense to track all the information related to our companies. A way to reduce this is to group these instructions into headings and use AI to summarize what they say.
Semrush AI’s new visibility tool has a feature called “Quick Research” that matches your keyword and title and provides you with a list of alerts alongside mentioned products, intent, and sources.
Currently, the tool allows you to filter results between the US and the UK, including full AI Feedback and product listings and URLs.
Even if I typed in a single keyword (“used cars”), it picked the closest available topic (“Used Car Sales and Sales”) and returned all the instructions, product mentions, and resource domains.
You may decide not to track a single piece of information, which can grow quickly and be difficult to measure. Instead, use the Semrush optimization database and measure the results by looking at the overall theme performance.
Without getting laid, your chances are slim
Remember that not all data needs a RAG, meaning that if the answer is already in the training data of the AI Search platform, no pages will appear as sources. For some brands, just talking about it is fine. If, say, someone is looking for a museum or a restaurant to visit, that mention may be enough to convince him to access the place and convert offline (eg, buy a ticket or a meal).
In most cases, however, SEOs are still looking for traffic, so the information must be written on their pages to provide it. opportunity to be seen. Ironically, while the results you get from ChatGPT are one answer instead of SERP, LLM actually does the background search for you.
Fortunately, you can find:
- Searches performed by ChatGPT in the background
- Chances are this search requires RAG
You can find this by looking for “queries,” “search_queries,” and “search_prob” within the Chrome Dev Tools (Inspect > Network > Discussion > Feedback).
Or, for convenience, you can add this script as a bookmark in Chrome and click on it after asking a question in ChatGPT. This is an improved version of Ziggy Shtrosberg’s script.
Although these searches look like traditional searches as opposed to commands, your strategy may be to optimize them and win over AI searches as a secondary benefit.
When it comes to search_prob (and the script above), the probability that the answer needs to be focused (RAG). This response is between 0 (low) and 1 (high). Every answer is different (even if you and I are searching for the same information, we will have different answers), so this can act as a proxy for the possibility of pages being listed as a source.
As with all new technologies, things change quickly. How people use AI tools and what tools are used is constantly changing. New models (such as ChatGPT5) are changing the way RAG is used, and the increase in adoption across different industries also affects the guidelines to follow, so you must also evolve and re-evaluate what and how to track AI searches.
See the complete picture of your search visibility.
Track, optimize, and win in Google search and AI from one platform.
Start a Free Trial
Start with
Contributing writers are invited to create content for Search Engine Land and are selected for their expertise and contribution to the search community. Our contributors work under the supervision of editorial staff and contributions are assessed for quality and relevance to our students. Search Engine Land is owned by Semrush. The contributor has not been asked to speak directly or indirectly about Semrush. The opinions they express are their own.
