Skip to content
Visbee Docs

Mentions & sentiment

The mention feed is where raw visibility data becomes actionable. Every AI response that Visbee captured during a collection appears here as a separate row. You can read the actual text an AI platform produced, see how your brand was framed, and trace which sources it cited. This is the evidence behind the summary numbers on your Dashboard.

What the mention feed shows

Section titled “What the mention feed shows”

Each row in the feed represents one AI response from one platform during one collection run. The columns you see at a glance are:

  • Platform. Which AI (ChatGPT, Claude, Perplexity, Grok, or DeepSeek) produced this response.
  • Date. When the collection that captured this response completed.
  • Snippet. A short excerpt from the AI answer, showing the section where your brand appeared.
  • Sentiment. A label (positive, neutral, or negative) indicating how the AI framed your brand in that response.
  • Brand. If you track multiple brands, this column identifies which one was mentioned.

Understanding sentiment

Section titled “Understanding sentiment”

Sentiment in Visbee is qualitative, not a number. Each mention is labelled as positive, neutral, or negative based on how the AI answer frames your brand in context.

A positive label means the AI described your brand favourably: recommending it, highlighting strengths, or framing it as a leading option for the use case being discussed.

A neutral label means the AI mentioned your brand in a factual or balanced way, without a clear recommendation or criticism.

A negative label means the AI raised concerns, compared your brand unfavourably, or otherwise framed it in a way that would not encourage a reader to choose it.

Sentiment is assessed per response, not per brand overall. A single brand can have positive mentions on one platform and neutral or negative mentions on another, which often points to specific content or coverage gaps worth addressing.

Filtering the feed

Section titled “Filtering the feed”

You can narrow the feed down using the filters at the top of the page. Available filters include:

  • Platform: see only responses from a specific AI.
  • Date range: focus on a specific collection period or compare two time windows.
  • Sentiment: isolate positive, neutral, or negative mentions to prioritise follow-up.
  • Brand: if you track more than one brand, filter to a single one.

Filters can be combined. For example, you might filter to negative sentiment on a single platform to find the specific AI responses where your brand is being framed poorly.

Expanding a mention

Section titled “Expanding a mention”

Clicking any row in the feed opens the full AI response. In the expanded view, your brand name is highlighted so you can immediately see how it was used in context. The sources cited by the AI in that response are listed below the answer text.

This is particularly useful when diagnosing a negative or neutral mention: you can read the exact language the AI used, see which competing brands appeared in the same response, and check which publications the AI drew on. Those sources are often the most direct lever you have for improving how the AI discusses your brand over time.

For a deeper look at which domains are cited most often across all your mentions, see Sources and domains.

Exporting mentions

Section titled “Exporting mentions”

You can export the current filtered view as a CSV file. The export includes all columns visible in the feed: platform, date, sentiment, brand, and the full answer text. This is useful for sharing findings with your team, building a report, or doing further analysis in a spreadsheet.