> ## Documentation Index > Fetch the complete documentation index at: https://docs.citino.ai/llms.txt > Use this file to discover all available pages before exploring further. # What we do not measure > The answers, engines, markets and signals outside Citino's sample, stated plainly so you can read every number for what it is. *Version 1 · 7 November 2026* Citino measures a sample: your questions, asked daily on the engines you track, from one market. Here is what falls outside it. ## Signed-in and personalised answers Every answer is collected signed out. A buyer who is signed in may see a different answer, shaped by their chat history, memory, saved preferences or location. We cannot see those answers, and no one outside that buyer's account can. ## More than one answer a day Each question gets one answer a day from each engine. We do not yet ask the same question several times a day, so the variation within a single day is not measured. The ± range accounts for this by being built across many questions and days. ## Other markets and languages Every project asks from the United States, in English. Other countries and languages are not available yet. A project has one market: measuring a second one would be a second project. The questions Citino asks about your brand for Objections are written in English and are asked only for English-language markets. ## Engines outside your sample * **Microsoft Copilot** is not tracked. Its answers can only be read signed in. * **Claude Sonnet, Grok and Claude Opus** are collected on Enterprise plans only. * **Gemini, Google AI Mode and Claude Haiku** are collected only for projects that switch them on. * **The Claude and Grok apps** are not read. Where Claude or Grok is in your sample, its answers come from the model's API, which is not the same page a buyer sees in the app. ## Visits we cannot attribute [Attribution](https://docs.citino.ai/guides/screens/attribution.md) uses your GA4 totals, never individual visitors, and is an estimate rather than a count. * Visits from Copilot, Claude and Grok are counted and labelled "not sampled", never dropped. * Clicks from Google's AI Overview and AI Mode arrive from the same address as ordinary Google search, so a visit cannot be told apart from an organic one. Attribution has no signal for those two surfaces. ## What an engine does not report * Not every engine says which pages it read. Shares of pages read are worked out over the answers whose engine reports it, and the screen says so. * When Google shows no AI Overview for a search, there is no answer to read, and it is not counted as an answer. ## Questions that name a brand Questions that name your brand or a competitor are collected every day and count toward your plan, but they are kept out of your headline numbers so they cannot inflate them. Their own figures are not shown yet: their row on Prompts and their Prompt detail read "needs one full run". ## A crawler visit is not a citation [Crawler activity](https://docs.citino.ai/guides/screens/crawler-activity.md) shows which AI crawlers fetched your pages. A fetch does not mean an answer used or cited the page. The two are shown side by side and never added together, and no alert is raised on crawler visits. ## Answer text through MCP The [MCP server](https://docs.citino.ai/mcp/data-handling.md) returns numbers, names and dates. It never returns an engine's answer text, because that text is written by a third party and an AI tool could read instructions inside it. The full text is on the **Answers** screen, and through the [API](https://docs.citino.ai/api/introduction.md) for a key that may read runs. ## Our question list We publish how questions are built, in [How answers are collected](https://docs.citino.ai/guides/methodology/collection.md), but never a list of questions. Publishing one would invite optimising for exact strings instead of for the questions buyers ask. ## Related > [card] [How answers are collected](https://docs.citino.ai/guides/methodology/collection.md) > > What we do collect, and how. > [card] [Engines and opt-ins](https://docs.citino.ai/guides/account/engines.md) > > Which engines your projects are asked on. > [card] [How data is handled](https://docs.citino.ai/mcp/data-handling.md) > > What AI tools connected to Citino can read. > [card] [FAQ](https://docs.citino.ai/guides/help/faq.md) > > Short answers to common questions. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.