Overview
Chats are the individual AI responses AI Brand Monitor collects when asking your prompts. Each chat represents one prompt sent to one AI platform on one day - giving you detailed insight into exactly what AI is saying about your brand.
What’s in a Chat?
Each chat contains:| Component | What It Includes |
|---|---|
| Prompt | The question that was asked |
| Platform | Which AI (ChatGPT, Perplexity, etc.) |
| Response | Full text of the AI’s answer |
| Brands Mentioned | Brands detected in the response |
| Sources Cited | URLs the AI referenced |
| Timestamp | When the response was captured |
| Sentiment | How positive/negative toward your brand |
Viewing Chats
Recent Chats Widget
- Recent - Latest AI responses
- Mentioned - Recent chats mentioning your brand
Chat Detail View
Click any chat to see the full response:Understanding Responses
Common Response Patterns
AI responses typically follow certain patterns:Recommendation Style
Recommendation Style
“Based on your needs, I’d recommend [Brand A] for X, [Brand B] for Y, and [Brand C] for Z.”Direct recommendations often mention multiple options.
Comparison Style
Comparison Style
“[Brand A] offers feature X while [Brand B] excels at Y. [Brand C] is best for budget-conscious users.”Helpful for understanding your competitive positioning.
List Style
List Style
“Top options include: 1. Brand A 2. Brand B 3. Brand C”Your position in these lists matters - being first or second is significantly better.
What to Look For
| Question | Why It Matters |
|---|---|
| Is your brand mentioned? | Basic visibility check |
| What position are you in? | First/second position gets the most attention |
| How are you described? | Reveals your perceived positioning |
| Who else is mentioned? | Shows competitive landscape |
| What sources are cited? | Identifies influential content |
| Is my persona applied? | See how different personas affect responses |
Brand Mentions in Chats
How Detection Works
AI Brand Monitor scans each response for:- Your brand name
- Your brand aliases
- Competitor brands
Mention Quality
| Mention Type | Example | Impact |
|---|---|---|
| Primary Recommendation | ”I recommend [Brand]“ | High positive |
| First in List | ”[Brand], X, Y are good options” | Positive |
| Among Options | ”Consider X, Y, [Brand], or Z” | Neutral |
| Comparison | ”[Brand] vs Competitor: Both are solid” | Neutral |
| Negative | ”Avoid [Brand] due to…” | Negative |
Sources in Chats
What Chats Reveal About Sources
Sources cited in chats show you:- Where AI gets information
- Which content influences recommendations
- What types of content AI trusts
Source Analysis
For each chat, you can see:- URLs cited
- Domains cited
- How many times each was referenced
Chat Types
Successful Chats
Normal responses containing:- Readable response text
- Optionally cited sources
- Brand mention analysis
Failed Chats
Sometimes AI requests fail due to platform timeout, temporary unavailability, or rate limiting. Failed chats are logged but don’t affect your metrics.
Executed Prompts (with Personas)
What’s an “Executed Prompt”?
| Field | Description |
|---|---|
| Original Prompt | What you created |
| Executed Prompt | What was actually sent (with persona applied) |
Best Practices
Regular Review
Daily Quick Check
Review the recent chats widget to spot any significant changes or concerning responses.
Weekly Deep Dive
Read through detailed responses for key prompts to understand nuanced positioning.