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

# A short glossary

> These definitions explain the terms you will meet while using IQRush.

These definitions explain the terms you will meet while using IQRush. The metric, selected run, and filters determine the scope of a result.

## Set up the measurement

| Term | Plain meaning |
| - | - |
| **Workspace** | A place to group projects for a brand, client, or team. |
| **Project** | The brand, website, topics, prompts, and engines set up to be measured together. |
| **Run** | One measurement of a project at a point in time. Completed means processing has finished; check stability separately. |
| **Baseline / reference** | The result used as the comparison point. Read the dates and label for the specific card or table. |
| **Topic** | A subject you want the project to measure. |
| **Prompt** | A question used in the measurement. |
| **Query** | A search phrase that can help identify topics and inform a generated prompt set. |
| **Intent** | The purpose of a question, such as learning or comparing options. |
| **Prompt type** | The kind of question, such as a comparison or a request for recommendations. |
| **Platform / AI engine** | The AI service represented in the selected results. |
| **Brand terms** | Names used to identify your brand, products, or services in the measurement. Review specific alternative spellings and acronyms; avoid generic category terms. |
| **Non-branded prompts** | Questions that let the engine choose which brands to introduce, without naming your brand or competitors in advance. |
| **Branded project** | A deliberately separate study using questions about a named brand or product. It measures that stated context, rather than unaided discovery. |

## Read visibility and sentiment

| Term | Plain meaning |
| - | - |
| **Response / answer** | An AI engine’s answer to a prompt. One response can contain several statements, citations, and mentions. |
| **Statement** | An individual passage or assertion within a response. It is not the unit counted by response coverage. |
| **Citation** | A reference to a source in an AI answer. |
| **Mention** | A recorded occurrence of a brand or another named entity in an answer. |
| **Entity** | A named thing, such as a brand, product, or ingredient. |
| **Domain** | A website address considered as a site, such as colorescience.com. |
| **URL** | The address of an individual page or resource. |
| **Citation share** | The website’s share of citation activity among the sources in the measured answers. A combined-platform result may differ from a percentage made from pooled raw counts. |
| **Citation coverage** | The percentage of measured responses that cite the domain’s pages at least once. |
| **Mention count** | The number of recorded mentions for the selected brand or entity; a response can contain more than one. |
| **Mention coverage** | The percentage of measured responses that mention the selected brand or entity at least once. |
| **Coverage** | Response-level presence: for one platform and selection, appearing in three of six responses gives 50% coverage. Each response counts once for the selected coverage measure. |
| **Grounded mentions** | Mentions linked to a website reference. The trend card reports their share among the analyzed mentions in the latest run. |
| **Net sentiment** | Positive share minus negative share among analyzed mentions; it describes framing, not visibility coverage. |
| **Positive / Negative / Neutral / Mixed** | Sentiment categories for analyzed mentions. Neutral is neither positive nor negative; mixed combines praise and criticism. |
| **Rank in field** | Position on the selected metric among the measured sources or entities. Tied values can share a position. |
| **Competitive gap** | Difference from the comparison named on the page, expressed in percentage points for the percentage metrics. |
| **Percentage points / pts** | The direct difference between percentages. Moving from 20% to 25% is an increase of five percentage points. |
| **Market concentration** | How much of the measured activity is held by the groups named on the card, not a measure of business revenue or website traffic. |

## Judge the evidence

| Term | Plain meaning |
| - | - |
| **Decision-grade** | Sufficiently relevant, stable, and precise evidence to support the decision being made. It is not guaranteed by a fixed prompt count or a completed status. |
| **Signal** | A measured pattern or difference that can be distinguished from uncertainty in the assessment. |
| **Noise** | Variation that can make a result or apparent change look different without enough evidence of a meaningful difference. |
| **Rank stability** | Whether the ordering of measured sources has settled as responses accumulate. Its percentage is not prediction accuracy. |
| **Signal-to-noise ratio (SNR)** | A measure of how clearly the measured pattern stands apart from measurement noise. It is a ratio, not a percentage. |
| **Domain SNR** | The SNR for the domain-level citation-ranking assessment. The Results stability card shows it against a target of 1.00. |
| **Results stability / Converged** | The assessment that both rank stability and signal sufficiency have passed for the citation-ranking result. Neither check alone is enough. |
| **Not converged** | The assessment has not passed both checks. Treat the ranking as provisional and investigate the available evidence. |
| **Confidence interval (CI)** | A range expressing uncertainty around a measured estimate. Wider intervals mean less precision. |
| **95% CI** | A confidence interval reported at the 95% confidence level. It describes uncertainty in the measurement; it is not a 95% promise that an AI answer is correct. |
| **Significant change** | A change the comparison identifies as distinguishable from measurement uncertainty. Its practical importance still needs judgment. |
| **No significant change** | The comparison does not establish a clear change beyond its uncertainty. It does not prove identical results. |

## Work with assistants

| Term | Plain meaning |
| - | - |
| **Agent Q** | IQRush’s in-platform assistant for exploring project results and working with recommendations and content. |
| **MCP / Model Context Protocol** | The connection standard used to give a supported assistant access to IQRush tools and authorized project data. |
| **Connected assistant / client** | An application such as ChatGPT or Claude connected to IQRush through MCP. Its available actions depend on the connection and permissions. |
| **Conversation** | A saved discussion with an assistant. Agent Q, ChatGPT, and Claude keep separate conversation histories. |
| **Context** | The project, run, filters, page, recommendation, or content a question refers to. Check attached chips and state the scope you mean. |
| **Audience profile / persona** | The audience perspective represented by a set of measured questions. A report needs relevant prompts or an available profile to make a measured audience comparison. |
| **PDP / product detail page** | A page describing an individual product, including its features and purchase information. |

## Work with recommendations and content

| Term | Plain meaning |
| - | - |
| **Recommendation** | A proposed action based on the available project results. |
| **Content Agent** | A group for a type of content work on the Recommendations page. |
| **Brief** | The audience, purpose, structure, and requirements for a piece of content. |
| **Content Playground** | The editor used to write, review, and score a draft. |
| **Segment / section** | A part of a page or draft that can be inspected separately. |
| **AI answers matched** | Measured answers that the content is relevant to in the stated topic or project context; it is not a future citation probability. |
| **Citation opportunity** | A relevant match identified as an opportunity, rather than an actual recorded citation. |
| **Strong / Moderate / Weak sections** | Groups describing the available citation and relevance evidence for sections. They are not editorial grades or a project-convergence verdict. |
| **Content readiness** | Checks for issues to review in a draft. These are separate from its matched-answer result and the project’s statistical stability. |
| **Autosaved** | The current working text has been saved. |
| **Version** | A saved copy of the document that can be inspected or restored. |
| **In review** | A writing status indicating readiness for review; it is not approval or publication. |
| **Implemented** | A recommendation whose completed action has been recorded. |

**Next:** [Understand your results](/guides/06-understand-results) or [start a task](/guides/14-choose-recommendations).


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