AI Visibility: Prompts Insights User Guide
The Prompts Insights module is the investigative heart of the AI Visibility suite. It moves beyond high-level scores to show you the exact questions users are asking and how different AI engines (LLMs) respond to them. By analyzing these prompts, you can understand the "reasoning" behind your brand's rankings and uncover specific gaps in your content strategy.
1. Navigating the Prompts Dashboard
The main dashboard serves as a searchable database of every query the system tracks for your brand.
Managing Your Queries
Search & Filter: Use the search bar to find specific prompts or filter the list by Themes to see how you perform in specific topical clusters.
Add Prompts: Click + Add Prompt to manually input specific queries you want to monitor (e.g., "Is [Brand Name] better than [Competitor]?").
Group by Theme: Toggle this view to organize prompts into topical "buckets," making it easier to see which product lines or service areas have the strongest AI presence.
Key Data Points
For every prompt in your list, the tool tracks these verified performance indicators:
Visibility %: The percentage of times your brand appears in the response for this specific prompt across all executions.
Position: Your brand's specific rank within the AI-generated answer.
Executions: The total number of times the tool has "asked" this prompt to gather data.
Citations: How often your website is explicitly referenced as a primary source.
Mention: A binary indicator showing if your brand was named in the text, even if a link wasn't provided.
Competitors: A quick-glance list of other brands appearing for the same query.
Last Updated: The exact date and time this prompt was last refreshed.
2. The Prompt Deep Dive (Expanded View)
Clicking on any individual prompt opens a detailed overlay for "forensic" analysis of AI behavior. This view is divided into specific performance tabs and granular data tables to help you understand exactly how LLMs evaluate your brand.
Visibility Tab
This tab provides a high-level view of your brand's presence for the selected prompt across various platforms.
Visibility Score: A line chart visualizing your brand’s presence over time for this specific query.
Visibility Score Rank: A bar chart comparing your brand’s rank across major AI Engines (AI Overviews, Perplexity, Claude, and ChatGPT).
Average Position Tab
This tab focuses on where your brand is placed within the AI’s response.
Average Position: A line chart tracking your numerical placement (e.g., #1 or #3) over time.
Average Position Rank: A bar chart showing how your placement consistency compares across different AI platforms.
The Responses Tab
This tab shows the raw, "live" data from every instance the AI was "asked" this prompt. It includes the following data columns:
Date: The exact time the execution occurred.
Platform: The specific AI Engine used (e.g., Google AI Overview, ChatGPT, Claude).
Region: The geographical location for the response (e.g., US).
Response: A snippet of the text generated by the AI.
Rank: Your brand’s specific placement within that individual answer.
Mentioned: A visual confirmation (checkmark) showing if your brand was present.
All Mentions: A list of icons for every brand (including competitors) named in that execution.
Citations: Icons representing the sources the AI used to build the response.
Fan Query: Displays the specific sub-queries the AI searched for to answer the user's prompt.
Response Detailed View
Clicking on any particular row in the Responses table opens an even more granular expanded view. This overlay displays:
Core Details: A summary of the Engine, Date, Visibility status, and Location.
Full Mentions & Fan Queries: A complete breakdown of entities detected and the AI’s reasoning path.
Full Response Text: The complete AI-generated answer to see the full context of your brand mention.
Source Citations: A card-based view of every URL, domain, and page title the AI referenced for that specific answer.
The Citations Tab
A dedicated view listing every source that the AI cited to verify its answer for this prompt.
Domain: The root website being used as a source (e.g., your site, a news outlet, or a review blog).
URL: The specific page link the AI referenced.
Title: The meta-title of the page, helping you identify which content types (e.g., "How-to guides" or "Product Features") the AI finds most trustworthy.
3. Advanced Reasoning: Query Fanout
Query Fanout represents a critical shift in AI search behavior. When a user asks a complex question, the AI doesn't just search for that one phrase; it "fans out" into multiple sub-queries to build a comprehensive answer.
How it works: If a user asks "Best CRM for small teams," the AI might fan out into "CRM with free tiers," "easy to set up CRM," and "HubSpot vs Zoho."
Why it matters: Seeing these sub-queries helps you understand the different "angles" the AI considers. If you are missing from a specific sub-query, you know exactly what kind of content to create to improve your overall ranking.
4. Sentiment Analysis: The Tone of Voice
AI doesn't just mention your brand; it evaluates it. The Sentiments section monitors how AI "feels" about your brand based on the following categories:
Positive Sentiment: Occurs when the AI recommends your brand, praises its features, or highlights its benefits as a solution to the user's prompt.
Neutral Sentiment: Occurs when the AI mentions your brand factually, such as in a list of providers or a technical description, without offering specific praise or criticism.
Negative Sentiment: Occurs when the AI highlights a drawback, mentions a limitation, or references negative user reviews and complaints found in its training data.
Occurrence Metric: This shows how often each sentiment type (Positive, Neutral, or Negative) appears across all executions of a prompt. High negative occurrence is a signal to update your public-facing content to address those specific concerns.
5. Citations & Attribution
Citations are the digital "footnotes" AI engines use to prove their answers are based on real-world facts. These are critical because they are the primary drivers of traffic back to your website.
Citation Share: The percentage of total citations you own within a specific category. (e.g., If there are 100 links in a theme and 13 are yours, your share is 13%).
Citation Rank: Your position on the leaderboard of cited domains. A rank of #1 means you are the most trusted source for that data set.
Citation Categories: A breakdown of source types, including Company Sites, Social (Reddit/YouTube), News, Reviews, and Tech.
Top Citation Domain: Identifies which specific website is the most "trusted" by the AI for your tracked prompts.
Top Citation Pages: A list of your specific URLs being cited the most, revealing which articles or videos provide the most value to AI engines.
6. Strategic Workflow
Use Prompts Insights to fuel your optimization strategy:
Identify High-Visibility / Low-Position Prompts: Find queries where you appear often but are ranked #4 or #5. These are the easiest to "bump up" with minor content tweaks.
Reverse Engineer the Fanout: Create new FAQs or articles that directly answer the sub-queries found in the Fanout section.
Dominate the Citation Share: If your visibility is high but your Citation Share is low, your brand is being mentioned but your site isn't being linked. Focus on building authoritative, "citeable" data assets.
Monitor Sentiment Spikes: If a previously "Positive" prompt turns "Neutral," check the Top Citation Pages to see if the AI has found a new, more critical source of information.
Bridge the Category Gap: Use Citation Categories to see if you are missing out on Social or Review site citations and adjust your presence on those third-party platforms.