AIRS Analyzer
The AIRS (AI Results) Analyzer lets you see how any query is answered across multiple AI platforms—from Google’s AI Overviews to ChatGPT, AI Mode, Gemini, Perplexity, Claude, Grok, and more. Understand where your brand appears and what sources AI tools are citing. For the strategic view of AI visibility monitoring, see AI Visibility and Competitors.
What You’ll Learn
Section titled “What You’ll Learn”- How to run an AIRS analysis
- Reading the Brand & Entity Landscape
- Understanding the comparison matrix
- Tracking brand mentions and citations
- Reading strategic insights
- Analyzing individual tool responses
- Managing past reports
- Exporting your findings
Getting Started
Section titled “Getting Started”What AIRS Reveals
Section titled “What AIRS Reveals”AIRS answers critical questions:
- Does your brand appear in AI-generated answers?
- Which sources are AI engines citing?
- How do AI answers compare to traditional SERPs?
- Where should you focus content efforts?
Running a New Analysis
Section titled “Running a New Analysis”- Navigate to AIRS Analyzer from the Toolbox
- Enter your search query (e.g., “best project management software”)
- Optionally select:
- Language — Target language for results
- Country — Geographic location the query is run from
- Brand — Select a brand to track mentions and citations of it throughout the report
- Click Analyze to start
Analysis Stages
Section titled “Analysis Stages”The analyzer runs through multiple stages:
| Stage | Description |
|---|---|
| 1. Fetching Answers | Queries all AI search tools simultaneously |
| 2. Analyzing Responses | Processes each tool’s response |
| 3. Comparing Results | Cross-references sources and citations |
| 4. Generating Insights | Creates strategic recommendations |
A full analysis usually takes several minutes, and complex queries can run 15 minutes or longer. Every AI engine is queried live before the analysis stages can begin, and the AI search platforms have been making their answers progressively harder to fetch, so those requests take longer than they used to.
You do not have to sit and watch it:
- Leave the page or refresh — the analysis keeps running, and returning to the AIRS Analyzer picks the progress display back up where it left off.
- We email you when the report is ready, so you can start one and come back later.
- Start another report while one is running — the first keeps processing in the background and appears in Past Reports when it finishes.
Understanding Your Results
Section titled “Understanding Your Results”Every report opens on three views. Switch between them at the top of the report:
| View | What it holds |
|---|---|
| Strategic Report | Strategic Summary, Brand & Entity Landscape, and Content Blueprints |
| Source Comparison | Which domains and URLs the engines cite, and how they overlap |
| Tool Results | Each engine’s full answer, its scores, and the entities it named |
Strategic Report
Section titled “Strategic Report”Strategic Summary
Section titled “Strategic Summary”- Overall AI Visibility — How often your brand/domain appears
- Key Findings — Most important takeaways
- Opportunities — Where you could improve visibility
- Threats — Competitive concerns to address
Entities and Brands
Section titled “Entities and Brands”Every entity the AI engines named in their answers, ranked by how many engines named it, with stance, sentiment and evidence quotes. Your selected brand is marked and pinned to the top of the section with its own coverage line. Full breakdown in Brand & Entity Landscape.
Content Blueprints
Section titled “Content Blueprints”A publishable brief per content opportunity: title, content goal, reader outcome, the role your brand plays in the piece, a full outline with H2 sections and their H3 subsections, the unique angle, key differentiators and the optimization strategy.
Source Comparison
Section titled “Source Comparison”Three things sit on this view.
Most-Cited URLs by AI Engines
Section titled “Most-Cited URLs by AI Engines”The URLs the AI engines cite most, ranked by how many distinct engines cite each one, with those engines named. Google and Bing organic results are excluded from this count. This table is about being quoted, not about ranking. Your own URLs are flagged with a You badge.
Citation Patterns & Opportunities
Section titled “Citation Patterns & Opportunities”Observed patterns in what the engines cite, each one written as three lines: the pattern, what it implies for your strategy, and the action to take. Each pattern lists the engines whose answers support it, so nothing here is an assertion you cannot trace back.
The Comparison Matrix
Section titled “The Comparison Matrix”Visibility across all platforms:
| Column | What It Shows |
|---|---|
| Google SERPs | Traditional organic rankings |
| Bing SERPs | Bing organic search positions |
| AI Mode | Google’s AI Mode responses |
| AIO | Google’s AI Overview citations |
| Bing Copilot | Microsoft’s AI assistant |
| Claude Web | Anthropic’s Claude with web search |
| Gemini | Google’s Gemini with grounding |
| Grok | xAI’s Grok with web search |
| OpenAI Web | OpenAI’s web search integration |
| Perplexity | Perplexity AI citations |
| ChatGPT | ChatGPT search results |
Reading the Matrix
Section titled “Reading the Matrix”- checkmark = Source appears in this tool
- Empty = Source not cited
- AI Citations = How many AI engines cite that source, shown as a count in the first column
- You badge = the source is on your brand’s domain
AI citations and organic search presence are counted separately and never merged. A page that ranks on both Google and Bing but is quoted by no AI engine has an AI Citations count of zero, which is the point. The table sorts by AI Citations (highest first) by default. Click any column heading to sort by that platform instead.
The Show sources cited by 3+ AI engines filter narrows the table to sources that more than one or two engines agree on.
Tool Results
Section titled “Tool Results”Deep-dive into individual engine responses:
- Full AI-generated text for each engine, with your brand mentions highlighted
- Tool performance matrix scoring each engine on depth, accuracy, intent match, and source quality
- Brand Mentions panel showing whether each engine named or cited you
- Response entities — a sortable table of every entity that engine named in its answer, with mention counts and stance
- Cited sources with links
- Error states if any engines failed
Brand & Entity Landscape
Section titled “Brand & Entity Landscape”The Brand & Entity Landscape answers one question: when AI engines answer this query, who do they actually talk about?
Floyi reads each engine’s answer, extracts every entity it names, then rolls those up into a single table across all responding engines.
The selected brand card
Section titled “The selected brand card”If you ran the analysis with a brand selected, the section opens with your brand: whether it was mentioned in the answers, in how many of the responding AI engines, and separately how many cited a page on your domain. Coverage is always stated against responding AI engines, meaning engines that returned a usable answer. Failed engines and the two SERPs are not in the denominator, so an outage cannot make your coverage look worse than it is.
The entity table
Section titled “The entity table”| Column | What it shows |
|---|---|
| Entity | The canonical name, with a Your Brand tag on your own |
| Type | Service provider, commercial brand, person, product or service, organization, directory or publisher, government, location |
| AI engines | How many of the responding engines named it |
| Total mentions | Mentions summed across every answer |
| Primary stance | Recommended, conditionally recommended, compared, mentioned, cited as authority, or criticized or discouraged |
| Sentiment | Positive, neutral, negative, mixed or unknown |
Sort by any column. Filter by entity type or by primary stance. Ten entities show by default, and View all expands the rest inline.
Per-entity evidence
Section titled “Per-entity evidence”Expand any row to see the case behind it: how many engines recommended it, the order it first appeared in, average sentiment confidence, any aliases the answers used, and one card per engine carrying that engine’s stance, sentiment, mention count and a quote from its answer.
Stance and sentiment are separate on purpose. An engine can name you in a neutral tone while recommending a competitor, and repetition is never treated as proof of quality.
Brand Mentions and Citations
Section titled “Brand Mentions and Citations”When you run an analysis with a brand selected, AIRS tracks two separate signals for every engine. They sound similar but mean very different things:
| Signal | What it means |
|---|---|
| Named | The engine wrote your brand name (or an alias, or your domain) into the answer itself |
| Cited | The engine used a page on your domain as a source, without necessarily naming you |
An engine can cite your page as a source and still credit the answer to someone else. That is a very different problem from not being in the answer at all, so AIRS never merges the two.
Where you’ll see it
Section titled “Where you’ll see it”- Selected brand card — at the top of the Brand & Entity Landscape: mentioned in X of Y responding AI engines, with citations stated separately underneath.
- Brand column — in the tool performance matrix on Tool Results, next to Overall. A filled badge means Named (with the mention count), an outlined badge means Cited only, and a dash means neither. Sort by it to see which engines know you.
- Report summary — a single line above the matrix: named in X of Y answers, cited by Z.
- Brand Mentions panel — for each engine, the mention count, citation count, which terms matched (brand name, aliases, or domain), and a quote showing the first mention in context.
- Highlighted responses — mentions are highlighted directly inside the AI answer so you can scan a long response and find them instantly. Use the Highlight in response checkbox to turn it off.
- Entity table — your brand is tagged Your Brand and shown alongside every other entity the engines named, so you can read your position relative to competitors instead of in isolation.
- Source Comparison table — rows on your own domain are marked with a You badge, so you can see at a glance which of your URLs the engines are pulling from.
How mentions are counted
Section titled “How mentions are counted”- The mention count reflects your brand in the answer text only. Sources are counted separately as citations, so a page title that happens to contain your name never inflates the mention count.
- Coverage is stated against responding AI engines, meaning the engines that returned a usable answer for this query. Engines that failed, engines disabled at the time of the run, and the two SERPs are excluded from the denominator.
- Matching covers your brand name, any aliases set in Topical Authority settings, and your domain (including variations like
www.and the bare domain name). This is the same matching Topical Authority uses, so the two tools agree on whether you were mentioned.
Strategic Insights
Section titled “Strategic Insights”AIRS provides actionable recommendations:
Visibility Scores
Section titled “Visibility Scores”- High Visibility — Appearing in 7+ tools
- Medium Visibility — Appearing in 4-6 tools
- Low Visibility — Appearing in 1-3 tools
- Not Visible — Not cited anywhere
Opportunity Areas
Section titled “Opportunity Areas”The analyzer identifies:
- Topics where competitors appear but you don’t
- Platforms where you’re underrepresented
- Content gaps to fill for better AI visibility
Managing Past Reports
Section titled “Managing Past Reports”Past Reports Tab
Section titled “Past Reports Tab”Access your analysis history:
- Search by query or date
- View previously analyzed topics
- Reload any past report
History Actions
Section titled “History Actions”- Load — Restore a previous analysis to view
- Delete — Remove reports you no longer need
- Bulk Delete — Select multiple reports to delete
Pagination
Section titled “Pagination”Reports are paginated. Click Load More to view additional history.
Regenerating Reports
Section titled “Regenerating Reports”Refresh Strategy
Section titled “Refresh Strategy”If results seem outdated or you want fresh analysis:
- Click Regenerate on the report, which shows its credit cost on the button
- Confirm in the Confirm Regenerate Strategy dialog
- The system re-analyzes the saved AI responses and generates new insights
Regenerating is also how you add the Brand & Entity Landscape to an older report. Floyi extracts the entities from the answers already stored with that report, so no engine is queried again.
Retry Failed Tools
Section titled “Retry Failed Tools”If any AI tools failed during initial analysis:
- Click Retry Failed Tools
- Only failed tools are re-queried
- Results are merged with existing data
Exporting Reports
Section titled “Exporting Reports”Export Formats
Section titled “Export Formats”Click the Export button on the report:
| Format | Best For |
|---|---|
| DOCX | Word documents, editing, white-labelling |
| Sharing, presentations | |
| HTML | Self-contained file for the web or email |
| Google Docs | Collaborating with your team |
The Source Comparison table exports separately as XLSX, CSV, or straight to Google Sheets, using its own Export button.
Export Contents
Section titled “Export Contents”Exports include:
- Strategic summary and insights
- Brand & Entity Landscape — the selected brand card and the full entity table with engine counts, mentions, stance and sentiment
- Content blueprints, with H2 sections and their H3 subsections
- Keyword & Query Signals — each keyword with the number of responding engines that used it
- Source Comparison — top domains, most-cited URLs, and the citation patterns with their recommended actions
- Tool performance matrix, including the Brand column
- Brand mention and citation status for each engine, with mention context
- Individual tool responses, with links, tables, and formatting preserved
- Source citations
The Source Comparison spreadsheet exports carry an AI Citation Engines column and an Organic Search Engines column, plus a Your Brand column marking rows on your own domain.
Frequently Asked Questions
Section titled “Frequently Asked Questions”What AI tools does AIRS query?
Section titled “What AI tools does AIRS query?”AIRS queries: Google SERPs, Bing SERPs, Google AI Mode, Google AI Overview, Bing Copilot, Claude Web Search, Gemini Grounding, Grok, OpenAI Web Search, Perplexity, and ChatGPT.
Why don’t some tools show results?
Section titled “Why don’t some tools show results?”Some AI tools may not return results for certain queries, or may experience temporary outages. Failed tools are marked and can be retried.
If a provider has an extended outage, we temporarily disable that engine so you are not charged for runs that cannot succeed. Disabled engines are hidden from new reports, while past reports keep the data they already collected.
Additionally, some AI search providers have geographic restrictions. For example, AI Mode and AI Overview are not available in Russia or Belarus, and ChatGPT has limited availability in certain regions. See Supported Countries & Regions for the full list.
How often do AI answers change?
Section titled “How often do AI answers change?”AI responses can change frequently. Run periodic analyses to track how your visibility evolves over time.
Can I analyze competitor visibility?
Section titled “Can I analyze competitor visibility?”Yes! Enter any query to see which domains appear. Compare your brand against competitors to identify gaps.
What’s the difference between this and traditional rank tracking?
Section titled “What’s the difference between this and traditional rank tracking?”Traditional rank tracking only shows organic SERP positions. AIRS shows where sources appear in AI-generated answers—an increasingly important visibility metric.
How are credits calculated?
Section titled “How are credits calculated?”AIRS uses a single bundle cost covering all 11 surfaces in one analysis: 9 AI engines plus Google and Bing SERPs. The exact cost depends on your AI model selection, and credits are only deducted when the analysis completes successfully.
What is the difference between being named and being cited?
Section titled “What is the difference between being named and being cited?”Being named means an engine wrote your brand into its answer. Being cited means it used one of your pages as a source without necessarily naming you. Engines regularly do the second without the first, which is why AIRS reports them separately. See Brand Mentions and Citations.
What is the difference between stance and sentiment?
Section titled “What is the difference between stance and sentiment?”Stance is how the answer positions an entity: recommended, conditionally recommended, compared against others, mentioned in passing, cited as an authority, or criticized. Sentiment is the tone toward it. An engine can describe you positively while recommending someone else, so the two are reported separately and neither is inferred from how often you are mentioned.
Why does an entity’s engine count look low?
Section titled “Why does an entity’s engine count look low?”Counts are stated against responding AI engines only, and each entity is counted from the answer text. If an engine failed on this run, it is not in the denominator. If an engine cited your page but never wrote your name, that is a citation, not a mention, and it appears in the Source Comparison view rather than the entity count.
Why does a URL show a high AI Citations count but no SERP position, or the reverse?
Section titled “Why does a URL show a high AI Citations count but no SERP position, or the reverse?”They measure different things and AIRS never merges them. AI Citations counts the AI engines that quoted a URL as a source. Organic search presence is whether it ranked on Google or Bing. A page can be quoted by six AI engines without ranking at all, and a top-three ranking page can go unquoted. The gap between the two columns is usually the most useful thing on the table.