# How Limy Generates Prompts

**Prompts** are the foundation of Limy. Everything in the platform starts with the questions users ask AI systems when researching your industry, products, and brand. 

Powered by **web intelligence**, these prompts act as the backend engine that drives all visibility measurement - allowing Limy to understand what your audience is asking, how AI systems respond, and how visible your brand is across large language models.

Prompts represent the questions users ask AI systems when researching products, companies, or topics in your market. In AI-driven discovery, users ask natural-language questions and receive a single synthesized answer — making whether and how your brand appears critical to visibility.

### **What Prompts Show**

Each prompt provides insight into:

* The latest AI-generated answer
* Brand and competitor mentions
* Brand sentiment
* Sources influencing the response
* Provider distribution

This helps you understand not just *if* your brand appears, but *how* it appears in AI answers.

### **Prompt Analysis**

The Prompts view displays all tracked prompts - both where your brand appears and where it doesn’t. You can search and filter by intent, topic, region, brand presence, or date range to identify high-impact opportunities.

Click [**Analyze**](#prompt-analysis) on any prompt to view detailed results, including competitors mentioned, sentiment, the AI's answer, and source attribution. Starring a prompt signals its importance and helps Limy prioritize recommendations.

<figure><img src="https://3636725193-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FsAix74DgbRDP7AprfL6m%2Fuploads%2FCK5YGPX73UW6lENGh05O%2FScreen%20Shot%202026-03-11%20at%2012.50.36%20PM.png?alt=media&#x26;token=48f7421b-3b70-4a8e-963c-e53c0f8f26e4" alt=""><figcaption></figcaption></figure>

<figure><img src="https://3636725193-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FsAix74DgbRDP7AprfL6m%2Fuploads%2FgibHYPzUqtrnFbmxabrn%2FScreen%20Shot%202026-03-11%20at%2012.51.32%20PM.png?alt=media&#x26;token=89c17265-f14b-468a-88a6-97162d135b3a" alt=""><figcaption></figcaption></figure>

### [**Modifying**](#modifying-prompts) **Prompts**

You can [add](https://support.limy.ai/intelligence/prompts-overview/adding-prompts), edit or remove prompts directly from the Prompts section to track custom questions, remove irrelevant ones, or reflect changes in your strategy. Updates apply to future prompt runs and visibility analysis.

**Starring a prompt** signals to Limy that this prompt should be prioritized when generating recommendations.

### Prompts are generated using a combination of:

* Your industry and category
* Your ICP and use cases
* Known AI search behavior patterns
* Competitive and category-level analysis

Limy continuously **maps the AI search landscape by identifying the types of questions users** **ask** when discovering, comparing, and evaluating solutions like yours.

These prompts reflect how users naturally interact with AI systems, rather than how they use traditional search engines. Limy originally expands each prompt into a set of related fan-out queries and **re-runs them every 24 hours to capture changes over time**.

### Fan-out Queries

When a user asks a question to an AI system, the system does not treat it as a single query. Instead, the AI internally explores multiple related variations and sub-questions to gather enough context to generate a reliable answer. These are known as **fan-out queries.**

Limy mirrors this behavior by generating clusters of related prompt variations, ensuring visibility analysis reflects how AI systems actually reason about your category, not just the surface-level question.

### Differences between Prompts vs. traditional SEO Queries

Prompts are fundamentally different from traditional SEO querie&#x73;**.** In classic search, users type short keywords and browse multiple results. In AI-driven discovery, users ask full, natural-language questions - *prompts -* and receive a single synthesized answer generated by an AI system.

While SEO queries optimize for ranking on a results page, prompts determine whether your brand is included, how it’s described, and which sources influence the AI’s response. This shift is why prompts and not keywords, are the unit of measurement for visibility in large language models.

### FAQs

<details>

<summary><strong>What is a prompt in Limy?</strong></summary>

A prompt represents a question a user might ask an AI system when researching a topic, product, or brand in your category.

</details>

<details>

<summary><strong>How are prompts created?</strong></summary>

Prompts are generated based on your industry, topics, target personas, and competitive landscape.

</details>

<details>

<summary><strong>What’s the difference between branded and non-branded prompts?</strong></summary>

Branded prompts explicitly mention your brand, while non-branded prompts focus on category or problem-based queries where your brand may or may not appear.

</details>

<details>

<summary><strong>Why do some prompts not mention my brand at all?</strong></summary>

AI systems reference brands based on perceived relevance, authority, and source coverage. Missing mentions highlight opportunities for improvement.<br>

</details>

<details>

<summary><strong>How often do prompts run?</strong></summary>

Prompts are run every 24-96 hours hours to get the latest accurate result.

</details>

***

### Related Content

{% content-ref url="prompts-overview/prompts" %}
[prompts](https://support.limy.ai/intelligence/prompts-overview/prompts)
{% endcontent-ref %}

{% content-ref url="prompts-overview/prompt-deep-analysis" %}
[prompt-deep-analysis](https://support.limy.ai/intelligence/prompts-overview/prompt-deep-analysis)
{% endcontent-ref %}

{% content-ref url="sentiment-score-analysis" %}
[sentiment-score-analysis](https://support.limy.ai/intelligence/sentiment-score-analysis)
{% endcontent-ref %}

{% content-ref url="../dashboard/dashboard-overview" %}
[dashboard-overview](https://support.limy.ai/dashboard/dashboard-overview)
{% endcontent-ref %}

Still have questions? We’re here to help. Contact us at **<answers@limy.ai>** - we’d be happy to assist.