In SEOmonitor the keyword you track is the topic you measure in AI Search, and its Google search volume is used as a proxy for that topic's demand. Because AI assistants don't expose prompt volumes of their own, this verified search volume is what puts AI Search on the same footing as your organic and AI Overview data, the same topics, weighted by the same demand. See more about this here.
To ensure the intent of the topic is carried correctly, SEOmonitor translates the exact keyword you track into a prompt. The platform then reports how that topic performs in AI Search: what the models answer, and, on commercial topics, which brands and sources they recommend.
This article explains how that prompt is built and how to read it in the product.
Note: This covers the prompt behind AI Search (LLM) tracking. AI Overviews (AIO/SGE) are triggered by Google on the keyword itself and are covered in their own article. For how AI Search results are scored (mentions, citations, sentiment, Share of Voice), see AI Search Tracking.
How we translate the prompt
Your keyword carries every intent Google answers for it. We read those intents from the live SERP (its features and top results), fold them into the one prompt a real person would ask an AI, and track that question. A few rules keep it faithful and consistent:
We cover the intents Google answers, dominant one first. A keyword often has more than one intent at once. For "paracetamol", Google answers both what the drug is and where to buy it, so the prompt covers both, leading with the dominant one. The dominant intent is the one shared by most of the top 10 results, so it is read from what actually ranks rather than assumed.
Products and brands carry a commercial intent. A specific product, model, or branded item almost always returns shopping and store results, so the prompt asks what it is, where to buy it, and the price, not an information-only version. For example, "dyson v15" becomes "What is the Dyson V15, how much does it cost and where can I buy it?"
Ambiguous keywords resolve to the meaning Google leads with. For a term with more than one meaning, we take the primary intent Google shows, preferring commercial over basic informational. "jaguar" with car results in the SERP becomes a question about the car, not the animal. A brand people navigate to becomes a navigational question.
We stay faithful to your keyword. We add no specifics your keyword does not contain (no places, prices, dates, models, or companies you did not include), and we keep your product or topic names exactly as written. We do not "correct" a name or spelling to a more famous lookalike, we translate the keyword you actually track.
We write in the language the campaign is set up for tracking. The question is phrased the way a real person asks an AI in that language, one natural question of a sentence or two, with accents and diacritics kept naturally.
How we track this prompt in LLMs
After translating the topic into a prompt, SEOmonitor runs it through the AI assistants you have enabled (ChatGPT, Gemini, Perplexity) in a fresh, neutral session with no history and no personalisation, localised to the location you have configured for your tracked site.
The same prompt is reused on every crawl, so your trend lines measure real change in the AI's answer, not a differently worded question each week.
How to find the prompt in SEOmonitor:
You can see this prompt in the keyword's AI Search snapshot, under How we ask AI → The prompt we track.
Below it, you see the outcome for that prompt: whether your site was named (mentioned), linked as a source (cited), or neither (not listed), and the same for your competitors.
For example, the keyword social media management tools (74K monthly searches) is tracked as:
"What are the best social media management tools, what is each best for, and how do I choose one for my business?"
That single question folds together the three intents the SERP shows for it: the comparison ("what are the best"), the use case ("what is each best for"), and the how-to-choose.
For the full definitions of Mentioned, Cited, Citation Rank, and Sentiment, and how they roll up into your scorecards and Share of Voice, see AI Search Tracking.
Good to know
The phrasing may differ from how you would ask it. The goal is not your exact wording, it is the question a typical person asks to get the answers Google gives for that keyword, phrased consistently so it stays reproducible across weeks and comparable across brands.
The prompt stays fixed. Once generated for a keyword, the prompt is stored and reused unchanged, so your week-over-week comparison reflects real movement in the AI's answer, not a reworded question.
AI answers are non-deterministic. Even with the same prompt, models can word answers differently between runs. That is normal, and it is why we focus on trends rather than single responses (see AI Search Tracking).