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

# Terminology

> Core terms and concepts used across the Topify.ai platform.

## Key terms

### Response

An individual AI-generated answer to a prompt. Also referred to as an **LLM response** or **chat**. Each response is collected from an AI provider (ChatGPT, Perplexity, Google AI Overview, Gemini, or Claude) and stored for analysis.

### Prompt

A question or query sent to AI platforms. Prompts represent the search queries your target audience might ask.

**Example**: "Which MOBA games are best for beginners?"

Topify.ai sends each prompt to multiple AI providers and collects their responses over time.

### Volume

Estimated AI search demand for a prompt -- how frequently people are likely asking similar questions to AI tools. Derived from keyword research signals, topic coherence, and intent type. This is **not** a count of responses collected by Topify.ai.

See [Volume](/concepts/brand-metrics#volume) for details.

### Visibility

A brand is **visible** when it appears in the list of brands extracted from a response. Visibility is a prerequisite for sentiment and position — if a brand isn't mentioned, those metrics aren't computed for that response.

### Position rank

Your brand's order among brands named in a response. For each response that mentions your brand, every brand is sorted by where it's first named (lowest character offset wins); your brand's rank in that response is its index + 1.

* Rank `1` = you're the first brand named
* Rank `3` = two brands are named before you in that response
* The dashboard shows the **average rank** across responses where your brand appears

<Tip>
  Character offsets are computed internally but not surfaced in the UI. What you see is the rank.
</Tip>

## Metrics at a glance

Topify.ai tracks five categories of metrics across your project.

| Metric                                                                                             | What it measures                                    | Scale                       |
| -------------------------------------------------------------------------------------------------- | --------------------------------------------------- | --------------------------- |
| [Visibility](/concepts/brand-metrics#visibility)                                                   | How often your brand appears in AI responses        | 0--100%                     |
| [Sentiment](/concepts/brand-metrics#sentiment)                                                     | How positively your brand is portrayed              | 0--100                      |
| [Position](/concepts/brand-metrics#position)                                                       | Where your brand appears relative to competitors    | Rank (1 = best)             |
| [Share of voice](/concepts/brand-metrics#share-of-voice-sov-your-slice-of-the-brand-mention-slots) | Your share of all brand-mention slots in AI answers | 0--100%                     |
| [Volume](/concepts/brand-metrics#volume)                                                           | Estimated AI search demand for a prompt             | Estimated monthly queries   |
| [Sources](/concepts/source-metrics)                                                                | How your content is cited by AI providers           | Citation count, %, averages |

## What "good" looks like

Each metric describes a different facet of your AI presence. None of them stands alone — they make sense together.

* **Visibility** is your share of voice inside AI answers. High visibility means your brand is consistently named when buyers ask AI about your category. **It is not the same as impressions** — visibility tells you *whether* you appear, not *how many people see* the answer.
* **Volume** is the demand side. It estimates how often real users actually ask each prompt. A prompt with high volume but low visibility is a high-leverage gap; a prompt with high visibility but near-zero volume is a vanity win.
* **Visibility × Volume** is the closest proxy to real AI-driven impressions. Optimize for prompts that score well on both.
* **Sentiment** only matters once visibility exists. A high sentiment with zero visibility means nothing — there were no mentions to score.
* **Position** only matters once visibility exists, for the same reason. Once you do appear, an earlier position usually translates to higher buyer recall.
* **Share of voice (SOV)** complements visibility by measuring how *crowded* the answers are. High visibility + low SOV means you appear often but in answers that name many brands; high visibility + high SOV means you dominate the conversation.
* **Sources** explain *why* your visibility is what it is. The websites AI cites most often for your prompts are the model's "trusted answers" — earning a place there is one of the most direct ways to lift visibility.

## Filters

All metrics can be filtered by **time period**, **AI provider**, and **topic tags**. For details on each filter and how they affect what's shown on the dashboard, see [Overview dashboard › Global filters](/guides/overview-dashboard#global-filters).