> ## Documentation Index
> Fetch the complete documentation index at: https://docs.fullreach.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# All answers

> Every collected answer, one row each, with filters on what an answer holds: the prompt, the brands it names and the sources it cites.

All answers lists every answer in the period, newest first. It is the place to read what the platforms said, one answer after the other. It also holds the full history of one prompt.

## The answer at the top

The page asks what AI answered when people asked these questions. The answer states how many answers the period holds, and how many of them name this brand. The line under it gives the same two counts for the newest day.

## The tiles

| Tile | What it counts |
| - | - |
| Answers | The answers in the period, with the prompts and the platforms they come from. An attempt with no answer does not count |
| Name this brand | Answers that name this brand. The share is the [visibility](/reference/metrics/visibility) of the period |
| Competitors only | Answers that name at least one monitored competitor and leave this brand out |
| No monitored brand | Answers that name neither this brand nor a monitored competitor. Such an answer can still cite sources |
| Cite this brand's site | Answers that cite a page on this brand's own website |

**Show these answers**, under a tile, filters the list to the answers that the tile counts.

## The list

Each row is one answer. A line above each day's rows states the day's answers and how many of them name this brand.

| Column | What it shows |
| - | - |
| Answer | The platform, the prompt, and the part of the answer that names a brand. This brand's name is marked in color |
| This brand | How the answer treats this brand: **Recommended**, **Named only**, **Advised against**, **Named** or **Absent**. See [Recommendation](/reference/metrics/recommendation) |
| Competitors | The monitored competitors that the answer names |
| Sources | The domains that the answer cites |
| Asked | The time of the collection. The header sorts the list newest first or oldest first |

**Named** is a mention that the analysis did not label. A row opens [the answer view](/reference/answer-view). **Previous** and **Next** there step through the list as it is filtered.

The list shows 25 answers on a page. The foot of the list states which answers the page holds and steps through the pages.

## The list's own filters

The search box finds a word in the prompts and in the answers. Five filters narrow the list:

| Filter | What it keeps |
| - | - |
| Prompt | The answers to one prompt, its earlier wordings included |
| This brand | Named, Recommended, Named only, Advised against or Absent |
| Competitor | One competitor, any competitor, no competitor, or no monitored brand |
| Source | This brand's own site, one of the most cited domains, or no source |
| Status | Answered, No answer, or both. Under No answer, **Failed** and **No AI Overview** each keep one kind of attempt |

Each option states how many answers it holds. **Clear filters** removes all of them.

These filters are in the page's address. So **Share** copies a link to the list as it is filtered, and other pages link to one view of it. The toolbar's filters apply too, but the toolbar offers no Prompt and no Competitor filter here. See [Filters](/reference/filters#where-each-filter-applies).

The tiles and the counts on the options read the toolbar's filters alone. A filter of the list does not change them.

## One prompt's full history

**See full history**, on the [All prompts](/reference/prompts) page, opens this list for one prompt. So does **All answers to this prompt**, in the answer view.

With a prompt picked, the list holds the answers to its earlier wordings too. Such a row carries the mark **Earlier wording**, with the day the wording changed. Those answers count in no figure of the All prompts page. The Overview counts them.

## Attempts with no answer

A platform can return no answer to a prompt. With **Status** on **No answer**, the list shows those attempts. Each row states the reason and says that it is not counted. When Google shows no AI Overview for a prompt, the row says so. That is a fact about Google, not a failure. An active prompt is asked again in the next daily collection.

## Where it appears

| Surface | Form |
| - | - |
| The All answers page | The list, its filters and the tiles |
| [The answer view](/reference/answer-view) | One answer in full |
| MCP, `list_prompt_runs` | The answers to one prompt |
| CSV export | One line per answer in the filtered list, with the text of the answer |


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.