Skip to main content
Prompts answers: which questions are we asking the engines for you, and which of them name you? One list holds every question in the project. Each question has its own page, Prompt detail, and new questions are added from the Add a prompt sheet.
The questions we ask the AI engines every day on your behalf, phrased the way a buyer would ask them. A question is never edited, only replaced, so its history stays comparable.

​
Prompts

​
The toolbar

  • The scope line, and a meter: how many prompts you are using against your plan’s limit. The limit is shared by every project in your workspace. See Plans and limits.
  • Add prompt opens the sheet described below.
  • Under the heading: asked daily on every engine; replace a question, never edit it.

​
Filters

Segment, Facet and Intent. The options come from your own questions, so a filter never offers a choice that matches nothing. If the three together match no question, the screen says so and offers Clear filters.

​
The table

A low-sample cue sits under a visibility figure that rests on only a few answers. Each row opens its Prompt detail.
A question that names your brand or a brand you track is asked every day and counts toward your plan, but it is kept out of your headline numbers so it cannot lift them. Its row and its Prompt detail show its own figures, which can read near 100% for a question with your name in it. Prompt detail says so beside the question: “Branded question · not in your headline”, or “Head-to-head question” when it names two or more brands. The row on this list carries no such label.

​
Prompt detail

One buyer question over time: who gets named, in what order, and which pages the engines read to decide.

​
The header

The question as it is asked, its status, its segments, its facets and the day it was first asked. Two controls sit beside it:
  • Answers per day opens a sheet with this question’s stored answers, day by day, each with whether you were named. Load earlier days goes further back.
  • Replace wording explains what a replacement does: it creates a new prompt whose history starts that day, and the current one stays readable, marked replaced, with no chart line drawn across the change. Creating the replacement from this screen is not available yet; it can be done through the API. A question that has already been replaced says so here instead.

​
What is on it, top to bottom

1

The verdict

Your visibility on this question with its n, and who was named instead (or as well). Beside it, one tile per engine: whether that engine named you, a link to its latest answer, and what the engine searched before answering. An engine that does not report its searches says so; one that reports them and ran none says “No searches recorded for this answer.”
2

Reachable sources

Of the answers whose engine told us what it read, the share that read at least one page you could get onto: anything not owned by a competitor. Why you lose this opens the question’s Source gap (see Gaps).
3

Who gets named

Every brand named in answers to this question, with the share of answers that name it and where it appears on average. Your own brand is highlighted.
4

Sources behind these answers

The sites these answers read and cited. Read is out of the answers whose engine reports what it read; Cited is out of every answer. A competitor’s own site reads “Competitor-owned · nothing to do”. All sources opens Sources.
5

Recurring searches

The searches the engines ran most often before answering this question, with how many answers ran each and which engines.
Every answer opens Answers, showing only this question.

​
Add a prompt

A sheet over Prompts, for a question your buyers ask that your list does not have yet.
  • Prompt text. Wording cannot be edited later: a question that changes is a different question, and its history starts again.
  • Intent. What the buyer is trying to do.
  • Market and Ask on are set for the project, so the sheet shows them without a control. Every engine the project asks is asked every question.
  • Segments. Only your claimed segments are offered, so the question shows up in your own figures.
  • The meter: the same count as the Prompts toolbar.
Add this question saves it, and it is first asked on the next daily collection. Cancel goes back to Prompts. If you already track that exact wording, or your plan’s prompt limit is full, the sheet says so under the question and keeps what you typed. Owners, admins and editors can add prompts. Viewers can read every prompt and its figures.

​
Empty and error states

  • You have no prompts yet: add the questions your buyers ask an AI assistant. Add a prompt opens the sheet.
  • No prompts match these filters: clear them to see the whole set.
  • No such prompt on Prompt detail: it may have been archived, or it belongs to another project.
  • No answers in this window yet: collection runs daily; the first answers land after the next run.

Answers

Read every stored answer in full.

Gaps

Why you lose a question, and what to do about it.

Scope and segments

Which questions count toward your scores.

Plans and limits

How many prompts your plan includes.