When your prospects ask ChatGPT or Gemini one thing, the mannequin quietly fires a set of conventional net searches within the background, retrieves the rating pages, and synthesizes the answer from those. The websites that rank for these hidden queries get cited. Those that don’t, don’t. QueryFan generates persona-specific prompts, runs them via each fashions, and captures the precise searches every one triggered. That listing is your actual AI visibility goal. It’s free.
Key phrases Lists Are Helpful, They Simply Miss Half The Image
Let me be exact about that earlier than anybody writes a livid reply.
I’m utilizing the time period “key phrases” to consult with the “one-shot” queries that go into conventional search engines like google and yahoo. Sure, I do know we’ve been in a “semantic” world for over a decade, however let’s simply agree on terminology that everybody can observe for now.
The first situation of “key phrase lists” in context to AI search is threefold:
- Sometimes, queries (prompts) that go into LLMs are typically longer, multifaceted, and conversational in nature. Conventional searches are typically extra slim in scope.
- Conventional search is “one-shot.” You do your search, get your info, then do one other impartial search. Queries/prompts on LLMs are typically conversational in nature and carry the context of earlier tokens.
- The mechanisms that LLMs use for net search additionally carry personalization context. If the consumer has beforehand acknowledged they’re a vegan, they usually ask the LLM about [running shoes], it’s extremely possible the LLM will carry out a search to accommodate this.
In essence, AI search has grow to be a sort of “common intent decoder” for customers. These huge, multifacted conversations with the AI get damaged down into subsets of solvable queries, that are run within the background as “conventional” searches on Google or Bing, with the ensuing websites used to generate a response. The method is named “Retrieval Augmented Era” (RAG).
The optimization goal has moved. You might be not optimizing purely for what the human varieties right into a chat field. You might be optimizing for what the AI agent quietly searches for on their behalf, within the background, with out the consumer understanding it occurred.
These background queries are what QueryFan captures. They’re typically fairly totally different from what the consumer really requested. And they’re the precise listing of issues that you must rank for to look in AI-generated solutions.
Exhibit A: Reddit Fell Off A Cliff On A Tuesday
The scope and depth of this secret relationship turned clear when Reddit was enjoying meteoric visibility increases in Google, and tragedy struck on September tenth, 2026. Based on quotation monitoring knowledge from PromptWatch, Reddit’s citation rate in ChatGPT responses collapsed virtually in a single day. It had been operating as excessive as 15% of all citations. Inside days, it was sitting beneath 2%.
The trigger was unglamorous: Google quietly removed the flexibility to request 100 search outcomes concurrently (the num=100 parameter) from its search API on that date.
Take into consideration what this tells you. Reddit’s visibility in ChatGPT responses tracked Google’s bulk search capabilities, not something Reddit did, not a coaching knowledge replace, not an alignment tweak. The implication is about as refined as a dropped piano: ChatGPT was bulk-pulling Google search outcomes, Reddit dominated these outcomes on the time, and when the bulk-pull disappeared, so did Reddit’s citations.
AI search surfaces are, in large part, wrappers around traditional search. The “AI” bit is actual (the synthesis, the personalisation, the conversational coherence) however the info retrieval step is remarkably acquainted. Google indexes and ranks the net; the AI consults that index. Your content material nonetheless must rank.
How QueryFan Works
Step 1: Your ‘Conventional’ Key phrases
Your conventional key phrase listing for the time period “trainers” could incorporate varied advised variations of this time period, from a supply like Google Counsel.
For QueryFan, we will merely take the subject of “trainers” and use this as our first step, as we’re going to generate prompts round this.
Step 2: Outline Personas
Your personas are how we’re going to customise the prompts we generate. This can alter our traversal of the token area, aligning us with coaching knowledge from the thousands and thousands of communities, discussion board posts, Reddit threads, and web discourse the place actual customers ask actual questions with these identities.
QueryFan sends your persona + matter mixture to the LLM to generate the sorts of questions that persona would really ask an AI device. Not key phrases. Questions. Actual, conversational, context-laden questions. For the [middle-aged vegan man who just started running] instance, it’s going to produce issues like:
- “Which vegan trainers are good for middle-aged males simply beginning to run?”
- “The place can I purchase vegan trainers on-line within the UK?”
- “What ought to I search for when selecting my first pair of trainers as a newbie?”
Step 3: LLM Choice And AlsoAsked Enrichment
AI conversations department. Somebody who asks about vegan trainers will ask follow-up questions: about price, about manufacturers, about damage prevention. QueryFan passes the generated prompts via the AlsoAsked API to seize the nearest-intent follow-up questions round every one. Folks Additionally Ask knowledge is the fitting instrument right here as a result of it was constructed to mannequin query proximity, which is exactly what you want while you’re making an attempt to foretell the place a dialog goes subsequent.
As an example, a search within the UK for “trainers” would floor observe up questions on particular manufacturers, asking methods to decide a shoe, and even widespread medical queries.
It’s also possible to choose if you happen to want to use ChatGPT, Gemini, or each. Every LLM handles and fan out queries barely otherwise, so if you happen to’re optimising for a particular platform it’s best to get the information from there.
Step 4: Question Fan-Out
QueryFan sends the enriched immediate listing to GPT-5 with net search enabled (through the OpenAI Responses API) and to Gemini with Google Search grounding lively (through the Gemini Grounding API). Each fashions, once they determine a immediate requires present info, carry out precise Google searches behind the scenes.
This course of captures the fan-out queries as each APIs are, moderately usefully, clear about what they searched. The Gemini API returns a webSearchQueries array within the groundingMetadata area of each grounded response. OpenAI’s Responses API logs the precise search queries within the web_search_call output. QueryFan harvests each.
The result’s a desk: persona-specific prompts in, the precise Google search queries the AI fired out. Not what your buyer typed. What the AI looked for on their behalf. These are your new search engine optimization targets, and till now there was no free device that surfaces them at scale.
The Grounding Query: Not Each Immediate Triggers A Search
A short however vital caveat earlier than you dash off to categorise every thing as an search engine optimization alternative.
Not every prompt causes the AI to perform a web search. The fashions decide primarily based on the consensus of token prediction as to if stay info is required.
To provide an instance, the immediate “What do crimson blood cells do?” doesn’t set off a search. The reason being there’s a very steep bell-curve of which tokens are going to look subsequent. Within the billions of coaching paperwork, the reply has stayed very secure, so an “in-model” reply can confidently be generated.
On the reverse finish of the size, a immediate resembling “What occurred within the information at this time?” would set off an internet search. There can be a really flat curve of “wtf tokens are subsequent?,” as there isn’t any “secure” reply inside the coaching knowledge; it all the time adjustments, it requires stay knowledge. It’s one other model of the Query Deserves Freshness (QDF) idea that SEOs have used for years.
For those who’re excited about grounding, Dan Petrovic has carried out some wonderful work on this space, and even released trained models on Hugging Face to foretell whether or not queries will likely be grounded once they hit a confidence threshold.
QueryFan surfaces which prompts triggered searches and which didn’t. Solely the grounded ones (those that truly prompted a Google search to occur) are actionable via search engine optimization. The in-model answers are, for now, largely outside your reach. You’d must affect coaching knowledge to maneuver the needle there, which is a distinct mission fully, with a for much longer horizon.
What You Do With The Outcomes
You now have a listing of precise search queries that AI instruments hearth when answering questions out of your particular personas. Run a typical hole evaluation:
- Which of those queries do you’ve content material for?
- Which do you already rank for?
- Which have zero protection, both in your web site or anyplace you’re more likely to be talked about?
The primary two classes are diagnostic. The third is your motion listing.
One vital distinction from conventional search engine optimization: Your own ranking isn’t the only path to AI visibility. LLMs scan the highest 10, 20, generally 50 outcomes for a grounded question and synthesize throughout them. A trusted evaluate web site rating at place 3 is a authentic path to showing in an AI-generated reply, even when your personal area by no means makes the primary web page. Getting a product reviewed on a high-authority specialist web site, incomes a point out in a roundup article, showing in related group content material, all of those rely.
LLM visibility is a multi-site focus. This implies the hole evaluation has two outputs: content material to create by yourself web site, and placements to earn on other people’s sites.
The Punchline
Solid your thoughts again to that Reddit quotation graph. The one which fell off a cliff when Google modified a single API parameter. A completely impartial firm’s AI visibility tracked the habits of a search API it didn’t management and possibly didn’t know existed.
That’s the form of the dependency. And the implication isn’t that search engine optimization is lifeless; it’s virtually the other. search engine optimization is now working at one extra take away: as an alternative of optimizing for the human question, that you must optimize for the AI-translated question that occurs between the human and Google.
QueryFan offers you a strategy to see what that translation really produces. Your key phrase listing tells you what folks typed right into a search bar. QueryFan tells you what ChatGPT and Gemini looked for on their behalf, within the background, with out anybody asking them to announce it.
These are totally different lists. The hole between them isn’t a minor refinement to your content material technique. It’s the a part of AI search that no person has been measuring as a result of no person has had a free device to measure it with.
Disclosure: The creator is the creator of Queryfan.
Extra Assets:
This put up was initially revealed on Mark Williams-Cook Substack.
Featured Picture: Roman Samborskyi/Shutterstock
Source link
