Using Gracker
Responses & Citations
Responses

Responses

Every time one of your prompts runs, GrackerAI captures the full AI response - the text, every brand mentioned, every source cited, and the sentiment. The Responses page is where you read what AI surfaces are actually saying.

Responses

If the Prompts page tells you what to ask, the Responses page tells you what AI actually answered. The most actionable insight lives here.

The Responses List

Each item in the list is one captured AI response: one prompt, one surface, one moment in time.

ColumnWhat It Shows
WhenWhen the response was captured
Answer EngineWhich AI platform produced this response
CountryThe country the prompt was run from
PromptThe prompt question that triggered this response

Click any row to open the full response.


Reading a Response

Opening a response shows a detail panel organised into tabs:

Response Details

  • Overview - the analysis scores for this single response (Visibility, Presence, Prominence, Sentiment, Citation, Mentions) plus response-quality facts like word count, tone, and format
  • Mentions - every brand the AI referenced, how prominently, and the source citations supporting each mention
  • Competitors - how each entity is positioned: role, factual accuracy, and citation support
  • Citations - every source domain cited, with URL and position
  • Attributes - descriptive labels the AI attached to each entity (e.g. "enterprise-grade", "complex to deploy")
  • Co-occur - pairs of entities that appeared together and how they were related
  • Response - the full AI-generated text exactly as produced, as rendered content

What 0% Visibility Means

A response showing 0% visibility means the AI answered the prompt but did not consider your brand relevant enough to mention.

This is your highest-priority signal. A 0% on a commercial-intent prompt where competitors appear means the AI sees your category but not you. That is a content gap, not a product gap.

Read 5–10 full responses on prompts where you score 0%. Look for patterns in which domains keep getting cited - those are the surfaces you need to compete on.

AI Surfaces

GrackerAI captures responses from 9 AI platforms:

ChatGPT
Broad recommendations, widely used by general buyers
Perplexity
Cites sources aggressively. Popular for research-heavy and technical queries
Google AI Mode
AI-first search experience inside Google
Google AI Overview
Quick inline summaries shown in classic Google search results
Microsoft Copilot
Dominant in enterprise and Microsoft-heavy organisations
Grok
Real-time search, growing among tech-forward audiences
Gemini
Strong on Google ecosystem topics. Pulls from YouTube and Reddit
Claude
Enterprise customers only
DeepSeek
Enterprise customers only

Filters

  • Prompt - show responses for a single prompt
  • Answer Engine - audit one AI platform at a time

To scope to a monitor group, use the monitor-group selector in the page header - it applies across the whole page, not just this table.


Exporting Responses

Click Export Excel in the top right to download your filtered response list as an XLSX file. The export includes only the rows currently visible after your filters are applied.

High-value export uses:

  • Content gap analysis: export 0% visibility responses on commercial prompts, hand to content team as a prioritised backlog
  • Competitor intelligence: export responses where a specific competitor appears
  • Sentiment audit: export low-sentiment responses, share with customer success or product team
  • Leadership reporting: quarterly visibility by surface

Common Workflows

Weekly Gap Analysis (15 min)

  1. Filter by Monitor Group to your category prompts
  2. Open responses where Visibility is 0%
  3. Read 5–10 responses and note which brands and domains are cited
  4. Hand patterns to your content team as the week's priorities

Competitor Intelligence Sweep (30 min)

  1. Filter by Monitor Group to a competitor group
  2. Read the 20 most recent responses
  3. Note how AI describes their differentiators, which sources are cited, which use cases they're winning
  4. Share findings with sales and product teams

Sentiment Audit (15 min, monthly)

  1. Open responses and look for low Sentiment scores in the panel
  2. Read the full response text for low-sentiment results
  3. Identify the content source feeding the negative framing
  4. Either request a correction or counter with new authoritative content

Next Steps