Open guides
Understand the retrieval layer.
Concise explanations for using AI-query evidence without overstating what the data can prove.
What are fan-out queries?
A practical explanation of the narrower searches answer engines can use to retrieve evidence before responding.
Why AEO and GEO need query data
How human demand, observed retrieval queries and estimated fan-outs support different AEO and GEO decisions.
Observed vs. estimated AI search queries
How Open Queries labels direct interface evidence and controlled model reconstructions without mixing them.
Estimating fan-out queries with log probabilities
A mathematical account of provider-native inverse perplexity, token alignment, repeated sampling and the limits of closed-model reconstruction.
Intent pillars
Apply the evidence to a concrete AI search job.
ChatGPT search queries
ChatGPT search queries are targeted web searches created when ChatGPT uses search to answer a request. Open Queries records only explicit search-tool query metadata exposed by the supported interface, never the prompt or conversation text, and labels any later fan-out reconstruction as an estimate.
Read the pillar →Clear terminologyChatGPT search history is not chat history
ChatGPT history is not public by default. Other people can view a conversation only when it is deliberately shared with them, subject to the link and workspace controls. Open Queries keeps a separate, local trace of surfaced web-search queries; it does not read prompts, responses, titles, account identity or conversation URLs.
Read the pillar →Supported surface · Claude web searchClaude web search queries
Claude web search queries are search strings issued when Claude uses its web-search capability. Open Queries records only explicit search-scoped tool metadata exposed by the supported interface and keeps model-generated fan-out candidates in a separate estimated evidence class.
Read the pillar →Supported surface · Google AI OverviewsGoogle AI Overviews: queries, fan-out and evidence
Google AI Overviews are AI-generated summaries in Google Search that include supporting web links when Google determines the feature adds value. Google says AI Overviews and AI Mode may use query fan-out—multiple related searches across subtopics and sources—to develop a response.
Read the pillar →AI retrieval vocabularyWhat are fan-out queries in AI search?
Fan-out queries are narrower searches created from one broader information need. An AI search system can issue multiple related queries across subtopics and sources, then use the retrieved evidence to assemble an answer. Google publicly describes query fan-out for AI Overviews and AI Mode.
Read the pillar →AI visibility · MeasurementAI search visibility starts with retrieval evidence
AI search visibility is the measurable presence of a source across retrieval, citations, referrals and outcomes in AI-assisted search. Open Queries contributes one narrow layer—query evidence. It is not a complete AI visibility tool, rank tracker or brand-monitoring suite.
Read the pillar →AI search optimization · Practical guideAI search optimization: an evidence-led workflow
AI search optimization is the practice of making useful information easy for search and answer systems to retrieve, understand and support with evidence. It combines ordinary technical SEO, answer structure, source quality and measurement; it does not require special AI-only markup or keyword repetition.
Read the pillar →Generative engine optimization · GEOGenerative engine optimization (GEO): a practical guide
Generative engine optimization is the practice of improving how accurately useful information can be retrieved, understood and cited in generative answers. Good GEO combines people-first content, technical search eligibility, source quality and honest measurement rather than relying on keyword repetition or guaranteed-citation claims.
Read the pillar →Answer engine optimization · AEOAnswer engine optimization (AEO): a practical guide
Answer engine optimization is the practice of structuring accurate, well-sourced information so people and answer systems can identify the answer, its scope and its evidence quickly. AEO improves answer clarity and retrievability; it does not guarantee a featured result or assistant citation.
Read the pillar →