TL;DR — AI builds a narrative about your brand every time someone asks, and you need to read it verbatim before you can manage it. Go to Visibility > Responses to read the exact paragraphs AI generates, then cross-reference with Strategy > Brand Perception to see which themes dominate across all responses. Check the Sentiment Analysis overlay to distinguish between being mentioned frequently and being mentioned favorably. Pro tip: a high visibility score only means AI talks about you a lot — always read the actual response text to know whether that attention is an asset or a liability.
The Question
“What does AI actually say about my brand when someone asks about it?”When a prospect asks ChatGPT, Claude, or Perplexity about your brand — whether to compare options, evaluate a purchase, or understand your positioning — they receive a generated paragraph, not a search result. That paragraph shapes perception before your website is ever visited. Knowing exactly what it says, and why, is the starting point for any AI brand strategy. This question covers the full content of those responses: the claims made, the attributes highlighted, the tone used, and the recurring narrative that emerges across models and prompts. You might also be wondering:
- “Is what AI says about me accurate, or is it distorted by outdated training data?”
- “Which attributes does AI emphasize most when describing my brand?”
- “Do all AI providers describe my brand the same way, or do they diverge?”
Where to Go in Qwairy
Start here: Visibility > Responses
Navigate to Visibility > Responses — this is where you read the actual text AI generated.
Focus on the response body for each answer: read the full paragraph, not just the score. Use the Provider filter to compare what ChatGPT says versus what Claude or Perplexity says for the same prompt.
Go deeper: Strategy > Brand Perception
Cross-reference with Strategy > Brand Perception to move from individual responses to aggregated signals.
Use the Theme extraction panel to identify which attributes appear most frequently across all responses. Use the Period filter to see whether the narrative has shifted recently.
Triangulate: Insights > Sentiment Analysis
Open Insights > Sentiment Analysis to overlay emotional tone on the content you just read.
A response can mention your brand frequently but in a cautious or comparative framing — Sentiment Analysis quantifies that distinction.
What to Look For
Responses — Verbatim AI Output
The Responses page lists every AI-generated answer collected during your monitoring runs. Each row shows the prompt, the provider, the date, the visibility score, and the full response text.| Element | What it tells you |
|---|---|
| Response body | The exact narrative AI constructs about your brand |
| Visibility score | How prominently your brand appears in the answer (0–100) |
| Provider badge | Which AI model generated this specific answer |
| Prompt context | The question that triggered this response — useful for understanding framing |
| Cited sources | The URLs AI referenced to construct its answer |
Brand Perception — Aggregated Narrative
Brand Perception processes all collected responses and extracts structured signals: dominant themes, attribute clusters, and a narrative summary that describes how AI collectively portrays your brand.| Element | What it tells you |
|---|---|
| Theme cloud | Which topics and attributes appear most often in AI descriptions of your brand |
| Narrative summary | A synthesized paragraph of what AI “believes” about your brand |
| Attribute strength | How confidently AI associates each attribute with your brand |
| Comparison baseline | How your narrative compares to competitors in the same space |
Pro Tip: Combine the verbatim Responses view with the Brand Perception theme summary. If a theme appears in Brand Perception but you cannot find it in actual responses, check the date range — it may reflect older data that has since shifted.
Filters That Help
| Filter | How to use it for this question |
|---|---|
| Provider | Isolate a single AI model to understand its specific narrative about you |
| Period | Compare what AI said last quarter versus this quarter to detect drift |
| Topic / Tag | Narrow to a product line, market segment, or use case to get targeted narrative analysis |
| Prompt | Read all responses generated by a specific monitoring prompt to see consistency |
How to Interpret the Results
Good result
AI responses consistently describe your brand using the attributes you intend to own: your category leadership, key differentiators, or core value proposition. The narrative summary in Brand Perception closely matches your actual positioning. Sentiment is predominantly positive or neutral-informative. Across 3+ providers, the descriptions are similar in substance, even if phrased differently.Needs attention
AI introduces attributes you do not claim, omits your primary differentiators, or describes your brand in a category you are trying to exit. Responses contain factual errors (wrong pricing, discontinued products, obsolete team details). Different providers give contradictory accounts — one positions you as enterprise-focused while another describes you as a startup tool.Example
Scenario: A luxury watch brand notices that AI responses consistently describe them as a “mid-range fashion brand” rather than a heritage watchmaker. Their recent repositioning toward haute horlogerie and limited-edition craftsmanship is not reflected in how ChatGPT, Claude, or Perplexity describe them. They want to understand the full narrative before deciding how to intervene.
- Open Visibility > Responses, filter by Provider = “All” and Period = “Last 90 days”. Read 10–15 responses containing the brand name. Note the recurring descriptors: “fashion-forward”, “accessible luxury”, “popular among millennials” — none of which reflect their heritage positioning.
- Open Strategy > Brand Perception and review the Theme extraction panel. Confirm that “fashion accessory” and “affordable luxury” are the dominant clusters. Note that “Swiss craftsmanship”, “in-house movement”, and “limited edition” — their core positioning pillars — do not appear.
- Open Insights > Sentiment Analysis and confirm that while sentiment is positive, the positive signals cluster around “style” and “trend” rather than “exclusivity” or “horological tradition”.
- Export the Brand Perception summary and bring it to the brand strategy team as evidence that AI is reinforcing the old mass-market narrative rather than the new heritage positioning, informing a content push toward watchmaking authority publications and collector communities.
Go Further
Export responses and perception data
Export AI-generated responses and brand perception data as CSV or XLSX for stakeholder presentations
Brand narrative tracking dashboard
Build a brand narrative tracking dashboard in Looker Studio using the answer-details data source
Share a perception snapshot with stakeholders
Create a shared view of your brand perception data for leadership or investors to review
Related Questions
Is the AI perception of my brand aligned with our actual positioning?
Compare what AI says versus what you intend to communicate — and measure the gap.
What are the strengths and weaknesses AI associates with my brand?
Extract a structured SWOT picture from AI-generated brand descriptions.
Do different AI providers describe my brand consistently?
Detect where ChatGPT, Claude, and Perplexity diverge in how they portray you.
Am I mentioned when users ask about my category?
Shift from what AI says to whether AI mentions you at all.