Skip to main content
Prompts are the queries that Mentionpath runs against AI platforms on your behalf. Every time someone asks an AI system a question that is relevant to your market, you want to know whether your brand appears in the answer — and how prominently. Prompts are how you tell Mentionpath which questions to watch. Each prompt is tracked across one or more AI providers (ChatGPT, Perplexity, Claude, Gemini, and others) and one or more geographic locations. Mentionpath runs each prompt on a recurring schedule, captures the AI response, detects brand mentions and citations, and feeds the results into your Analytics and Opportunities dashboards.

Creating prompts

You can create prompts in three ways. Click Create prompts in the top-right corner of the Prompts view to access all three options.
1

Create manually

Type a single prompt and press save. Use this for specific, high-priority queries you have already identified. Set the location and assign topics during creation.
2

Create with AI

Describe a theme or category and let Mentionpath suggest a batch of prompts. You review the generated list and accept the ones you want. This is the fastest way to build a comprehensive prompt library from scratch.
3

Import from CSV

Upload a CSV file with your existing prompt list. Useful when migrating from another tracking tool or when you have a pre-defined keyword list from SEO research.

Prompt types

Every prompt has a type that describes its search intent. The type affects how Mentionpath scores and categorizes the prompt, and how results are interpreted in Analytics.

Category

General queries about a product category or topic area. Example: “best project management software for remote teams.”

Comparison

Queries that compare options or ask for alternatives. Example: “Notion vs Monday.com vs Asana.”

Use case

Queries tied to a specific problem or scenario. Example: “how to track OKRs for a startup.”

Brand intent

Queries that reference your brand directly. Example: “does Acme integrate with Salesforce.”

Integration

Queries about integrations, connections, or ecosystem fit. Example: “tools that integrate with HubSpot.”

Prompt scoring

Each prompt receives four scores that help you prioritize where to focus your efforts. Scores are computed automatically after prompts are created and updated periodically.
ScoreWhat it measures
VolumeEstimated search demand for this query (1–5 bars). Higher = more people are asking this question.
DifficultyHow competitive the prompt is across AI responses (1–5 bars). Lower = easier for your brand to appear.
AI likelihoodHow likely AI platforms are to generate a substantive response for this query, rather than deferring to a link list.
OpportunityA composite score combining the above factors to indicate overall priority.
Scores appear as colored bar indicators in the prompts table. Volume bars are green when high (high opportunity) and red when low. Difficulty bars are green when low (easy) and red when high (competitive). While a prompt is being scored, the bars display a loading animation.
Newly created prompts show scoring as pending until Mentionpath has processed them. This typically takes a few minutes after creation.

Locations and geographies

Each prompt can be tracked across multiple geographic locations. Click the location chip on any prompt row (or the Add Location chip if none is set) to open the location editor. You can assign multiple locations to a single prompt. Mentionpath runs the prompt separately for each location, since AI responses can vary significantly by region. For example, “best accounting software” may return different top brands in the United States versus the United Kingdom. When creating prompts via AI generation, Mentionpath defaults the location to your brand’s primary location from your Domain DNA profile. You can change this during the creation flow. To bulk-assign locations to multiple prompts at once, select them using the checkboxes and click Add Locations.

Managing active and archived prompts

The status dropdown in the toolbar filters between Active and Archived prompts.
  • Active prompts are tracked on an ongoing basis. They consume tracking capacity and appear in Analytics.
  • Archived prompts are paused. Historical data is preserved, but no new runs are executed.
To archive one or more prompts, select them and click Archive. To restore archived prompts, switch to the Archived filter, select the prompts, and click Unarchive.
Archive prompts that are no longer strategically relevant rather than deleting them. This keeps your historical data intact for trend analysis.

Viewing prompt detail

Click any prompt row to open its detail view. This view shows everything Mentionpath has learned about how a specific query performs across AI platforms. Metrics strip — Five headline numbers at the top: Mentions (runs where your brand appeared), Sources (citations from your domain), Average Position, Volume indicator, and Difficulty indicator. Sentiment distribution — A color-coded bar and score rings showing how positive, neutral, or negative AI responses are for your brand in this prompt. Includes separate scores for overall sentiment, recommendation strength, and salience (how central your brand is in the response). Performance by Platform — A tabbed chart showing Mentions, Citations, or Position broken down by AI provider (ChatGPT, Perplexity, Claude, etc.). This tells you which platforms are most favorable to your brand for this specific query. Brand Rankings — A table of all brands that appear in AI responses for this prompt, ranked by mentions, with sources, average position, and share of voice. Click Manage Competitors to update which brands are tracked as competitors. Response History — The full text of recent AI responses for this prompt, with your brand mentions highlighted. Filter by platform or location. Scroll to load older responses.
The Response History section renders AI responses with citation links intact. Numbers in brackets like [1] link to the original sources the AI cited.

Filtering and searching

The main Prompts table supports several filtering options:
  • Search — Full-text search across prompt titles (debounced as you type).
  • Status — Toggle between Active and Archived.
  • Topics — Click the Topics column header to filter by one or more topic labels.
  • Locations — Click the Locations column header to filter by specific geographies.
  • Sort — Click the Volume or Difficulty column headers to sort by those scores.
To bulk-edit topics or locations across multiple prompts, select them with the checkboxes and use the Add Topics or Add Locations buttons that appear in the toolbar.