> ## 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. # Sources > Which websites the engines read while answering your questions, which of them you can do something about, and what the engines searched for. Sources answers: which pages do the engines actually read before they decide who to name, and which of those could you get onto? A switch at the top moves between two tabs: **Domains**, the sites read, and **Searches**, the web searches the engines ran first. Each site opens its own **Source detail**. Sources always covers the last 30 days. It has no window control and does not follow the window picked on Overview. ## Two shares, two denominators Most rows on this screen carry two shares. They are worked out over different sets of answers, so they are never compared with each other or added together. * **Retrieved**: "The share of answers that read a page from this site while writing the answer, out of the answers whose engine tells us what it read." Not every engine reports what it read, so this share leaves out the answers that cannot tell us. * **Cited**: "The share of answers that showed this site as a source the reader can click. A site can be read without being cited." This one is out of every answer in scope. Each share is printed beside its own count, never under the other’s heading. The summary blocks at the top count reads rather than shares, and never turn reads into a percentage: an answer that read three sites counts once for each, so reads add up to more than answers. ## Domains ### What is on it, top to bottom **1. What the engines read** Every website the AI engines read while answering your questions, each sized by how many answers read it. Beside it, **answers reading a source you can influence**: the share of answers that read at least one site you could do something about, such as a review site, a list, a directory, a community, a publication or your own site. Competitors’ own sites are not counted. **2. Where to show up first** The sites you can do something about that the engines read most. Getting onto these is the most direct way to be named more often. They are ranked by how often each is read, not yet by what being named there would recover. **3. Your site** Where your own website ranks among every site the engines read, by how many answers read it: how many answers read it and cited it, how many competitor sites are read more, and your most-read page. **See your pages** jumps to the per-page table. **4. Sources read** The full table, in three groups: sites **you can act on**, **competitor-owned** sites, and **yours**. Each row: its place by reads, the site and its type, **Retrieved** (the count and its share, with a bar), **Cited**, **Prompts** (how many of your questions had at least one answer that read this site) and **Your presence**. **5. Your pages** Every page of your own site an answer read: how often it was retrieved and cited, how many prompts, **First retrieved** (the first day any answer read the page, ever, not only in the last 30 days) and **Last seen**. ### Your presence Whether you are listed on a site is not checked yet, so a site you could act on reads **Unknown**. A competitor’s own site reads **Competitor-owned · nothing to do** and carries no action at all: being read there is not a way in for you. The **Only prompts I lose** switch is shown and not available yet. ## Searches Before an engine answers, it usually searches the web, and those searches decide which pages it reads. The Searches tab shows them, counted in this project’s answers. * **A coverage line** names the engines in scope that report their searches and those that do not. Not every engine reports them, so the list is never presented as complete. * **What the searches show**: how many of the searches name a brand you track, how many are new this period, whether the engines searched for your brand by name, the brands they looked up, and the words they searched most. * **Searches | Terms** switches the table. **Searches** lists each search word for word, with how many answers ran it, on how many prompts and which engines, and a **new** mark for a search not run in the previous window of the same length. **Terms** lists each word in those searches, with how many answers and how many distinct searches held it. These are plain counts with no `±`: a count of answers is not a sampled estimate. ## Source detail One site, in full. Open it from any row. * **The header**: the site, its type, your listing, and when the engines first and most recently read it. **All sources** returns to the list. * **Four figures**: **Retrieved** and **Cited** (the two shares above), **Prompts it is read on**, and **Of those, you are not named on**: of your questions whose answers read this site, how many have not named you once in this period. * **Prompts where it is read**: your questions whose answers read this site, with your visibility on each and the question’s intent. Each opens its Prompt detail. * **Brands it names**: the brands named in answers that read this site, and in how many of those answers, including whether you are. * **Your pages**, for your own site only: the same per-page table as above, plus **Fetches** and **Last fetched**: how often AI crawlers fetched the page, from your crawler logs. They sit beside the reads and are never divided by them. Without logs they read "logs not connected". See [Crawler activity](https://docs.citino.ai/guides/screens/crawler-activity.md). ## Empty and error states * **No source was read in this scope**: no answer in the last 30 days cited a page we could resolve. Widen the segment scope. * **No engine in this scope reported which sites it read**: it needs one full run. * **No such source**: none of this project’s answers have cited that site. * **Cited before, not in this window**: your answers cited the site before the last 30 days. * On Searches: **no engine in this scope reports its searches** (the engines are named), **no answers in this scope yet**, or **no searches in this scope**. ## Related > [card] [Gaps](https://docs.citino.ai/guides/screens/gaps.md) > > Turn a site you are missing from into a work item. > [card] [Answers](https://docs.citino.ai/guides/screens/answers.md) > > The answers that read each site. > [card] [Crawlability](https://docs.citino.ai/guides/screens/crawlability.md) > > Whether the engines’ crawlers can read your site at all. > [card] [Glossary](https://docs.citino.ai/guides/glossary.md) > > Every term the screen’s question marks explain. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.