AI Conversation Visibility
The AI Conversations Visibility module ranks monitored questions by the visibility of Your Brands across AI platforms. Use Best first to find questions where the brand appears most consistently and Worst first to find gaps to investigate.
Open Conversations
Open AI Visibility Dashboard, select a project and choose the Conversations tab in the table section. The project needs processed monitoring responses and at least one saved name under Your Brands.
Choose the ordering
- Best first shows Top Conversations by Visibility.
- Worst first shows Worst Conversations by Visibility.
The two groups are kept separate. When more conversations match than can be displayed, the note below the table tells you how many of the top or bottom conversations are shown.
What visibility means
For each monitored question, visibility is:
latest eligible responses mentioning Your Brands ÷ latest eligible monitored responses × 100
The table uses the newest processed response for each platform and monitoring date. Several aliases from Your Brands in one answer still count as one brand-visible response.
The current table counts any Your Brands appearance, including an appearance
without a stored position. It does not apply the global Provider Position
control.
Read a row
Each row identifies the monitored question and shows:
- the overall visibility percentage;
- the number of brand-visible and eligible responses; and
- results for each measured AI platform.
A platform value of 0% means eligible responses existed but none contained a qualifying brand appearance. A dash means that platform had no eligible observation under the selected filters.
Apply the dashboard filters
Use Date Range to choose the reporting period. Journey Phase narrows the
questions to Awareness, Consideration or Decision. Configuration isolates
one audience and question set. Platform restricts the table to one
platform. Provider Position is not applied by this module.
Archived or changed questions can still contribute to historical results when they have processed responses in the selected period.
Use the results
Start with Worst first and look for a question with enough eligible responses to be meaningful. Compare its platform values, then open the underlying AI conversation responses to read what providers were named and how the answer was framed.
Use Best first to identify themes where the brand already has strong visibility. Compare those questions with weak ones to find differences in audience intent, journey phase, evidence and competitor positioning.
What this module does not show
Visibility does not tell you whether the mention was positive, whether the brand beat a competitor, or whether a source cited your website. Use Conversation Sentiment Analysis, Competitive Win Rate and the citation modules for those questions.
If something does not look right
- Open Configure and confirm the names under Your Brands.
- Broaden Date Range, Journey Phase, Configuration and Platform.
- A 0% result is measured absence; a dash means no eligible observation.
- Settings changes can change historical percentages without changing the stored answers.
- If only one conversation matches, it can appear in Best while the separate Worst group remains empty.