Watch “How to read your overview in Peec AI” on YouTube
Your Overview dashboard shows key areas of your AI search performance, in addition to all recent chats generated for your prompt set. Use the filters at the top to focus your analysis.
Here you’ll see:
- Visibility graph: See where you stand against competitors across all your prompts, with daily fluctuations.
- Brands: Your average visibility, sentiment, and position compared with top competitors.
- Top sources: See which websites and domains AI platforms use most when answering your prompts, along with a breakdown of the source types they rely on, such as Corporate, Editorial, UGC, and more.
- Recent chats: See the latest AI responses across all your tracked prompts. By default, all recent responses are shown, but you can toggle the view to see which brands were mentioned alongside yours.
Dashboard filters
Use the filters at the top of your dashboard to narrow down your data and focus your analysis on what matters most.- Competitor: Filter by a competitor to see their visibility and which sources AI uses when answering about them.
- Date range: View data for different time periods (default is the last 7 days). Select a predefined range or a custom span. Change indicators to compare performance with the previous span of the same length.
- Tags: Filter by prompt tags using AND/OR logic. Select multiple tags to control which prompts are included:
- AND: Shows only prompts that have all the selected tags.
- OR: Shows prompts that have at least one of the selected tags.
- Models: Focus on specific AI platforms (ChatGPT, Claude, Gemini). Use this to see how you perform on individual models versus the average.
- Country: Filter by geographic location where prompts were run from (based on your prompt setup).
- Topics: Filter by specific topic folders to focus your analysis on particular prompt groups.
Tag filters don’t apply to the recent All Chats section; you’ll always see the recent chats regardless of tag selections.
The Country filter becomes available once you have prompts that are assigned to different countries.The Tag and Topic filters become available once you’ve assigned tags or topics to your prompts.
Saved filters
Save a filter combination you come back to, so you can apply it again in one click. A saved filter works on every page with a filter bar, and filters a page doesn’t offer are ignored on that page.- Saving: Once you’ve changed the filters, click Save and give the combination a name. With a saved filter applied, Save changes updates it and Save as new keeps it and saves your changes as a new one.
- Switching: After your first save, a dropdown appears at the start of the filter bar. It lists Default, the filters the page loads with, followed by your saved filters. Selecting one applies it.
- Undoing changes: Reset returns to the saved filter you last applied, or to Default if none is applied.
- Renaming and deleting: Hover over a saved filter in the dropdown and use the pencil or trash icon.
Saved filters are personal. They’re stored for your account in this project, so your colleagues don’t see yours. Anyone opening a shared dashboard link can’t save filters.
Main Overview
Visibility graph
The visibility graph shows your brand and your top 6 competitors, and how their visibility changes daily. This gives you a clear picture of competitive dynamics and helps you identify who’s gaining or losing ground in AI responses. Each line represents a brand’s Visibility % over your selected time period. Hover over a line to see the exact percentage and which brand or competitor it represents.Brands
This section shows the same 7 brands from the visibility graph, with their average performance over your selected time span:- Visibility: Percentage of AI responses where your brand appears. Shows how frequently AI models mention your brand when answering relevant prompts. A higher visibility percentage means better brand awareness in AI search.
- Share of Voice (SoV): Percentage of your brand mentions in AI responses compared to all tracked brands mentioned. A high SoV means that you’re more likely to be the main focus of conversation in chats than competitors.
- Sentiment: How positively AI platforms describe your brand (0 to 100 scale), based on the language and context around your mentions (from words like “trusted,” “reliable,” and “leading” to critical or negative associations). Higher scores mean more positive perception.
- Position: Your brand’s average ranking when it appears in AI responses. It shows where your brand ranks against all other brands mentioned, including both tracked and untracked competitors. Lower numbers are better: Position 1 means your brand is mentioned first, Position 5 means fifth, and so on.
Read more about identifying and adding competitor brands here.
Top sources
The Top Sources section shows which websites AI platforms reference most when answering your prompts, split by Source Type and Domain:- Source type chart: Shows the number of retrievals by domain category (Editorial, Corporate, UGC, etc.). This helps you understand what types of sources dominate in your industry.
- Top domains table: Lists the 8 most-used domains with their metrics:
- Retrieved: Percentage of chats where at least one URL from this domain was used as a source.
- Retrieval rate: Average number of times a URL from this domain was used as a source per chat.
- Citation rate: Average number of times the domain was explicitly cited when used as a source.
- Type: Domain category classification.
Links
The Links section shows the websites and pages AI answers link to in their own text, separate from the sources they cite. See Links for how a link differs from a source.- Linked domains: Domains ranked by the number of chats that link to them.
- Linked URLs: The individual pages, ranked the same way.
- Top: The domains or pages the most chats linked to in the selected period.
- New: Domains or pages linked to for the first time in the selected period.
- Trending: The largest rise in chats linking to them.
- Losing: The largest drop in chats linking to them.
All chats
These are the chats Peec AI creates by running your prompts daily against the AI models. By default, you’ll see all recent chats; you can use the filters at the top to filter by brand, features, and sources. If your visibility is zero for certain prompts, you can still see that the prompts are running and what the AI models are saying about your topic area. Each chat shows the prompt that generated it, followed by these columns:- Mentions: The tracked brands mentioned in the answer.
- Sources: The domains the AI used as sources.
- Links: The first three domains the answer linked to. Hover to see every linked domain and how many links went to each.
- Features: The chat features in the answer, such as ads or a web search.
- Citations: How many times the answer cited its sources.
- Position: Your brand’s position, if it’s mentioned.
- Created: When the chat ran.
Prompts
Navigate to All Prompts in your sidebar to see all the prompts you’re tracking. They are grouped by Topics (or under No Topic folder if none assigned). At the top of the page, you’ll see how many active prompts you have versus your plan limit. Prompts are organized into three tabs:- Active: Prompts Peec AI runs daily across the AI models.
- Suggested: Generated prompt suggestions with Volume data, suggestion date, and Reject/Track buttons.
- Archived: Paused prompts that you can reactivate at any time or permanently delete along with all their collected data.
- Visibility: Percentage of chats mentioning the brand in the selected time period.
- SoV: The brand’s mentions divided by all brand mentions across your prompts in the selected time period.
- Position: The brand’s average position when mentioned in the selected time period.
- Win rate: Percentage of chats where the brand is ranked first in the selected time period.
- Sentiment: The brand’s sentiment score when mentioned in the selected time period.
- Mentions: Brands mentioned in answers to this prompt or topic.
- Volume: Estimated search volume for a prompt relative to your industry. Hover over the colored bar to see the volume level, ranging from Very low to Very high.
- Web Search: Share of chats where the AI ran a live web search.
- Links: Share of chats whose answer links to a page in its own text, with the prompt’s top linked domains beside it. Hover over the domains to see each one’s share of the prompt’s chats.
- Branding: Whether the prompt mentions your brand. See Branded and Intent classification tags.
- Intent: The prompt’s search intent: informational, commercial, or transactional.
- Fact-checking: Whether claims from this prompt are checked against your facts. See Fact-checking.
- Tags: Tags you’ve added to organize prompts.
- Location: The geographic location the prompt runs from.
- Added: When you created the prompt.
Use the column settings to hide or reorder columns, or to add the chat features that aren’t shown by default, such as Ads or Maps. If you’ve saved your own column layout, it stays as you left it.
Individual prompt dashboards
Click any active prompt to see focused analytics for that prompt. You’ll see the same layout as your main Overview, filtered to just that prompt. Key differences:- No tag filters: You’re already viewing a specific prompt.
- Top 7 brands: How your brand is performing compared to your tracked competitors for this prompt. If they’re not in the top 7, they’ll appear at the bottom of the table with their actual ranking (e.g., #36).
- Query fanouts (only available for ChatGPT, Perplexity, and Copilot): The latest tracked fanout queries, so you can see what similar queries the model ran while creating the answer. For a more granular view, open the Fanouts tab in your sidebar.
- Common terms: The terms used most often across the fanout queries.
- Links: The domains and pages this prompt’s answers link to, in the same cards as the Links section on your Overview.
- Recent chats in table format: Shows recent responses in a structured table.
Query fanouts
To see a detailed breakdown of fanout queries, open the Fanouts tab in the sidebar.
What the page shows
AI models often run background searches before generating a response. The two stats at the top summarize those searches for the selected time period.- Distinct fanout queries: The number of unique fanout queries generated for the selected filters.
- Total occurrences: The total number of times those fanout queries were issued.
- Top term: Word pair with the most occurrences across all query fanouts in the selected period, ignoring the Terms filter.
Term share
Term share shows the percentage of query fanouts that contain a specific term, helping you understand how frequently AI models use certain terms when searching for information. You can explore this data in two views:- Over time: Track how frequently each term appears in query fanouts over your selected time span. Switch between daily, weekly, or monthly views to spot changes and trends.
- By model: Compare how frequently different AI models use each term in their searches.
Terms
This section shows the most common word pairs found in query fanouts, helping you understand what AI models are searching for when generating responses. You can explore terms using four views:- Top: The most frequently used word pairs.
- Trending: Word pairs that are increasing in occurrences.
- Losing: Word pairs that are decreasing in occurrences.
- New: Word pairs that have recently started appearing in query fanouts.
Brands
The Brands section shows which brands AI models mention in their query fanouts. The Occurrences column shows how many times each brand appears in these searches, helping you identify which brands AI models are searching for most frequently. This gives you a clearer picture of which competitors are appearing in AI searches and how often they’re being considered during the search process. Below the stats, the All queries list shows every fanout Peec captured, broken down by the topics you are tracking. Each row carries three things:- AI Model: The AI model that triggered the fanout query.
- Query: The exact background search the AI ran.
- Type: What kind of fanout it was: “Search, Shopping, or Synthetic”.
- Occurrences: How many times that query fired across your chats within the selected time period.
- Topics (default): Queries rolled up under each Topic, with a count beside it (for example, “Reputation questions” with 1,557). This is the fastest way to see what AI searches for in each area you track.
- Prompts: The same queries organized by the Prompt that triggered them, for when you want to trace fanouts back to a specific question.
Brand visibility vs source visibility
Understanding the difference between these two is important:- Brand visibility: Your brand is explicitly mentioned in the response generated by the AI model.
- Source visibility: Your domain or content was used as a source or cited, even if your brand name was not explicitly mentioned.
- If you’re cited often but never mentioned, it might mean your brand lacks authority or name recognition.
- If you’re mentioned often but never cited, AI might associate your brand name with a topic, but not trust your content as a reference.
