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

# Prompt Research

> Test prompts across AI platforms before committing a tracked slot to them.

Tracked prompt slots are limited and every GEO run bills for all of them. Prompt Research lets you trial candidate prompts first, see how the platforms answer, and only then promote the ones worth tracking.

## How it works

<Steps>
  <Step title="Pick or add a topic">
    The left column shows **Your topics** (from your [brand profile](/workspace/brand-profile)) and **Suggested topics**. **Generate more** appends fresh suggestions, and you can add your own.
  </Step>

  <Step title="Build a basket of prompts">
    Add prompts to test — either from the suggestions for a topic, or by writing your own. Each prompt is tagged with a topic and an intent category. Duplicates are rejected.
  </Step>

  <Step title="Run">
    Up to **10 prompts** per run, at **20 credits per prompt**. The cost is shown before you run.
  </Step>

  <Step title="Read the results">
    Each prompt returns the platform answers, whether you were mentioned or cited, and which competitors appeared.
  </Step>

  <Step title="Import the winners">
    Promote prompts worth monitoring into your tracked set. They're included from the next GEO run onward.
  </Step>
</Steps>

## Reading results

Results use the same vocabulary as [Prompts](/discoverability/ai-visibility/prompts) — mentioned, cited, competitor-only, neither — so what you learn here transfers directly.

Open a result for the full answer, its sources, and the competitors named. A run detail dialog holds everything for a completed run.

## Run history

Every run is kept with its status — **In progress**, **Completed**, **Failed**. Open a past run to re-read its results without paying to run it again.

## Import limits

Add-on prompt slots are fixed per site and set by your plan. When you're at the cap, importing requires removing an existing add-on prompt or upgrading.

<Card title="Plans" icon="credit-card" href="/account/plans">
  Prompt caps by plan.
</Card>

## What to test

<Tip>
  Good candidates: comparison prompts naming you and a competitor; buyer-intent prompts with pricing or procurement language; locational prompts if you sell regionally; and the exact phrasing your sales team hears on calls.
</Tip>

Bad candidates: prompts so broad that every answer is a listicle, and prompts that only your existing customers would type.

## Availability

Prompt Research is a paid-plan feature. Without it, the page shows an upgrade prompt.

<CardGroup cols={2}>
  <Card title="AI Visibility" icon="sparkles" href="/discoverability/ai-visibility/overview">
    Where promoted prompts end up.
  </Card>

  <Card title="Credits" icon="coins" href="/reference/credits">
    What a research run costs.
  </Card>
</CardGroup>