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

# Prompts & Topics

> Manage the queries PromptAlpha uses to monitor your brand

## What Are Prompts?

Prompts are the specific queries that PromptAlpha sends to AI search engines to monitor your brand. Each prompt simulates a real user question — the kind of query your target audience might type into ChatGPT, Perplexity, or any other AI engine. By tracking how AI engines respond to these prompts, PromptAlpha measures your [visibility](/docs/answer-engine-insights/visibility-scoring), [citations](/docs/answer-engine-insights/citation-mapping), [sentiment](/docs/answer-engine-insights/sentiment-analysis), and [competitive positioning](/docs/answer-engine-insights/competitor-benchmarks).

The quality and breadth of your prompt set directly determines the accuracy and usefulness of your Answer Engine Insights data. A well-curated prompt library ensures you are monitoring the queries that matter most to your business.

## Auto-Generated Prompts During Onboarding

When you set up PromptAlpha, the platform automatically generates an initial set of prompts based on:

* **Your brand name and domain** — prompts that directly reference your brand.
* **Your industry and category** — prompts reflecting common questions in your vertical.
* **Your product and service descriptions** — prompts targeting the solutions you offer.
* **Competitor landscape** — prompts where competitors are likely to appear.

<Info>
  Auto-generated prompts provide a strong starting point, but they should not be your only prompts. Review and supplement them with custom prompts that reflect your specific audience's language and intent.
</Info>

## Adding Custom Prompts

You can add custom prompts at any time from the **Prompts** tab in Answer Engine Insights.

<Steps>
  <Step title="Navigate to the Prompts Tab">
    Open **Answer Engine Insights** and select the **Prompts** tab from the navigation.
  </Step>

  <Step title="Click Add Prompt">
    Click the **Add Prompt** button. A form will appear where you can enter your custom query.
  </Step>

  <Step title="Write Your Prompt">
    Enter the query exactly as a user would type it into an AI engine. Write in natural language — AI engines respond to conversational queries, not keyword strings.
  </Step>

  <Step title="Assign a Topic (Optional)">
    Select an existing topic or create a new one to group this prompt with related queries. Topics help you organize and analyze prompts thematically.
  </Step>

  <Step title="Select Target Engines">
    By default, prompts are sent to all supported AI engines. You can optionally limit a prompt to specific platforms if it is only relevant to certain engines.
  </Step>

  <Step title="Save and Activate">
    Click **Save**. The prompt will be included in the next data collection cycle and results will begin appearing within 24 hours.
  </Step>
</Steps>

<Tip>
  Write prompts the way real users ask questions. Instead of "best CRM software," try "What is the best CRM software for small businesses?" AI engines respond differently to conversational queries than they do to keyword-style searches.
</Tip>

## Prompt Types

PromptAlpha categorizes prompts into three types based on user intent. Understanding these types helps you build a balanced prompt library that covers the full customer journey.

### Navigational Prompts

Navigational prompts are queries where the user is looking for a specific brand or product. These prompts include your brand name directly.

| Example                     | Purpose                                      |
| --------------------------- | -------------------------------------------- |
| "What does \[Brand] do?"    | Monitors how AI engines describe your brand  |
| "Is \[Brand] legit?"        | Tracks trust-related sentiment               |
| "\[Brand] vs \[Competitor]" | Monitors direct comparison responses         |
| "\[Brand] pricing"          | Checks how AI engines represent your pricing |

Navigational prompts reveal how AI engines perceive and describe your brand when users ask about you specifically.

### Informational Prompts

Informational prompts are queries where the user seeks knowledge about a topic related to your industry. Your brand may or may not be mentioned in the response.

| Example                                 | Purpose                                                   |
| --------------------------------------- | --------------------------------------------------------- |
| "How does \[category] work?"            | Monitors whether your brand is referenced as an authority |
| "What are the benefits of \[solution]?" | Tracks if AI engines cite your content as a source        |
| "Best practices for \[industry topic]"  | Measures topical authority and visibility                 |

Informational prompts measure your brand's authority and thought leadership in AI search.

### Commercial Prompts

Commercial prompts are queries where the user is evaluating options and likely preparing to make a decision. These are high-intent queries where being mentioned can directly impact revenue.

| Example                                | Purpose                                           |
| -------------------------------------- | ------------------------------------------------- |
| "Best \[product category] in 2025"     | Monitors whether your brand is recommended        |
| "Top \[solution type] for \[use case]" | Tracks inclusion in recommendation lists          |
| "\[Product category] comparison"       | Monitors competitive positioning in buying guides |

<Warning>
  If your prompt library is heavily weighted toward one prompt type, your data will have blind spots. A balanced mix of navigational, informational, and commercial prompts gives you the most complete picture of your AI search presence.
</Warning>

## Topic Grouping

Topics let you organize prompts into thematic groups, making it easier to analyze performance by subject area rather than reviewing individual prompts one by one.

### How Topics Work

* Each prompt can be assigned to one topic.
* Topics are user-defined — you create names that make sense for your business.
* The **Topics** view on your dashboard aggregates visibility, sentiment, and citation data at the topic level.

### Common Topic Structures

<AccordionGroup>
  <Accordion title="By Product Line">
    Group prompts by the product or service they relate to. This lets you compare AI visibility across your product portfolio.

    * "Enterprise Platform" — prompts about your enterprise offering
    * "Starter Plan" — prompts about your entry-level product
    * "API & Integrations" — prompts about technical capabilities
  </Accordion>

  <Accordion title="By Customer Journey Stage">
    Group prompts by where they fall in the buyer's journey.

    * "Awareness" — informational prompts about your category
    * "Consideration" — comparison and evaluation prompts
    * "Decision" — pricing, reviews, and recommendation prompts
  </Accordion>

  <Accordion title="By Industry Vertical">
    If you serve multiple industries, group prompts by vertical.

    * "Healthcare" — prompts specific to healthcare use cases
    * "Financial Services" — prompts targeting finance buyers
    * "E-commerce" — prompts about retail and e-commerce solutions
  </Accordion>
</AccordionGroup>

To create or manage topics, navigate to **Prompts > Topics** from the Answer Engine Insights dashboard.

## Best Practices for Prompt Selection

Building an effective prompt library is one of the most impactful things you can do to get value from Answer Engine Insights. Follow these guidelines:

<Steps>
  <Step title="Start with Customer Language">
    Review your customer support tickets, sales call transcripts, and search console queries to understand the exact language your audience uses. Build prompts that mirror real questions, not internal jargon.
  </Step>

  <Step title="Cover the Full Funnel">
    Include a mix of navigational, informational, and commercial prompts. Aim for roughly equal coverage across all three types to avoid data bias.
  </Step>

  <Step title="Include Competitor Comparisons">
    Add prompts like "\[Your Brand] vs \[Competitor]" and "\[Competitor] alternatives." These reveal how AI engines position you relative to specific competitors and are critical for [competitive benchmarking](/docs/answer-engine-insights/competitor-benchmarks).
  </Step>

  <Step title="Use Natural Language">
    Write prompts as complete, conversational questions. AI engines are optimized for natural language, and their responses to keyword-style queries may differ significantly from what real users see.
  </Step>

  <Step title="Review and Refresh Quarterly">
    Your market, product, and audience evolve over time. Audit your prompt library at least once per quarter to add new relevant queries, remove outdated ones, and adjust to shifts in customer language or competitive dynamics.
  </Step>

  <Step title="Monitor Prompt Performance">
    Use the prompt-level data in your dashboard to identify prompts that consistently return no brand mentions. These may indicate queries where you have a visibility gap — or they may be poorly phrased prompts that need rewording.
  </Step>
</Steps>

<Note>
  The number of prompts you can monitor depends on your PromptAlpha plan. If you are approaching your limit, prioritize commercial and navigational prompts, as these have the most direct impact on business outcomes.
</Note>