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

# Prompt discovery

> Browse and select from AI-generated prompt suggestions so you can expand what Erlin tracks without brainstorming from scratch.

Prompt discovery is Erlin's library of suggested prompts, generated from your brand, category, and competitive landscape. These are prompts you have not started tracking yet. Select the relevant ones, execute them, and they move into your active tracked set automatically.

## What prompt discovery does

When you set up your brand, Erlin analyzes your domain, industry, and positioning to generate a pool of relevant prompts. These are questions your audience is likely asking AI about your category. Prompt discovery surfaces them in a grid, grouped into clusters, so you can browse and pick the ones that matter most.

The clusters are organized by three dimensions:

* **Revenue role**: prompts grouped by the business outcome they target (lead generation, retention, upsell)
* **Funnel stage**: prompts grouped by where the user is in their journey (awareness, consideration, decision)
* **Content type**: prompts grouped by the kind of response they trigger (comparison, how-to, recommendation)

Prompts in the Library are not yet tracked. They do not run on scans and they do not feed into your dashboard. They only move into your active tracked set when you choose to execute them.

## How to use it

<Steps>
  <Step title="Open the Library tab">
    Navigate to **AI Visibility > Prompts > Library** (labelled "Prompt Discovery" in Erlin).
  </Step>

  <Step title="Browse clusters">
    Scroll through the prompt groups organized by revenue role, funnel stage, or content type. Each cluster shows a title and the number of available prompts inside.
  </Step>

  <Step title="Select prompts">
    Check the prompts you want to start tracking. You can select across multiple clusters in a single session. Prompts you have already tracked are hidden from the grid by default.
  </Step>

  <Step title="Assign a folder (optional)">
    Pick an existing folder from the dropdown to organize the new prompts, or let Erlin place them in the default folder.
  </Step>

  <Step title="Execute">
    Click the execute button. Erlin runs the selected prompts on every platform your plan enables (ChatGPT, Perplexity, Gemini, Claude, and Google's AI Overview and AI Mode), stores the results, and moves the prompts into the Tracking tab. Your AI Visibility dashboard updates automatically once the run completes.
  </Step>
</Steps>

A progress banner appears while execution runs. Results appear in your Tracking tab as each platform completes. After execution, the prompts disappear from the Library grid (since they are now tracked) and start contributing to your dashboard metrics on every future scan.

## Generating more prompts

If you have already tracked most of the initial suggestions and want fresh ones, click **Generate More** in the Library tab header. Erlin creates 10-20 additional prompts based on your brand profile, competitive landscape, and keyword data.

The generated prompts appear in the grid immediately, ready to select and execute. Generation typically takes 5-10 seconds. Erlin uses your brand analysis data (domain, products, services, competitors) to target gaps in your current tracked set and avoids duplicating prompts you already track.

## Languages

The language selector in the page header determines which language Erlin generates prompts in. Change it before clicking Generate More to produce prompts in a different language.

This is useful for brands that serve audiences in multiple markets. Generate English prompts, track them, then switch to Spanish and generate another batch for the same product in a different market.

## Why discovery is separate from tracking

Discovery is about expanding coverage: casting a wider net across questions your audience asks AI. Tracking is the active, measured set that runs on every scan and costs credits.

Separating the two means you can explore new prompt angles without increasing scan costs, review prompts before committing them to your active set, and generate in bulk but only promote the most relevant ones.

Once you are confident a prompt belongs in your dashboard, one click moves it from discovery to tracking. If a prompt turns out to be irrelevant after a scan or two, you can remove it from tracking without affecting your other data.

## Related

<CardGroup cols={2}>
  <Card title="Prompt tracking" icon="message-circle" href="/ai-visibility/prompts/prompt-tracking">
    Manage the active set of prompts that run on every scan.
  </Card>

  <Card title="AI Visibility" icon="eye" href="/ai-visibility/overview">
    See how tracked prompts roll up into your dashboard metrics.
  </Card>

  <Card title="Visibility Score" icon="message-circle" href="/ai-visibility/metrics/visibility-score">
    The primary metric that tracked prompts feed into.
  </Card>
</CardGroup>
