TL;DR — Open Analytics > Page Performance to see a composite AI score for every tracked URL, combining citation rate, crawler visits, referrer traffic, and GSC ranking into one number. Go to Visibility > Citation Sources to check citation diversity per page — a high score from one provider only is fragile. Cross-reference with Strategy > GSC to find pages that rank well in traditional search but are not yet cited by AI. Pro tip: sort by citation rate (not overall score) to find pages trusted by AI but not driving traffic — often a quick citation link format fix.
The Question
“What is the unified AI performance score for my most important pages?”Individual metrics — how often a page is cited, how frequently AI crawlers visit it, whether it drives referrer traffic — are each meaningful, but they do not answer the question a content team or CMO actually asks: “Is this page performing well in AI search overall?” The Page Performance AI score in Qwairy is a composite metric that combines all of these signals into a single number per URL, making it easy to rank your pages, spot the underperformers, and demonstrate content ROI. You might also be wondering:
- “Which pages on my site have the strongest AI presence?”
- “How is the AI performance score calculated and what does it mean?”
- “Which pages are being crawled by AI bots but not generating any citations?”
Where to Go in Qwairy
Start here: Analytics > Page Performance
Navigate to Analytics > Page Performance — this is your per-URL AI performance hub.
The default view lists all tracked URLs ranked by their AI performance score. The score is a composite of four signals: citation rate (how often the page is cited as a source in AI-generated answers), crawler visit frequency (how often AI bots are returning to re-evaluate the page), referrer-driven traffic (sessions arriving from AI tool referrers directly to this URL), and GSC organic ranking (how the page ranks in traditional search as a proxy for authority).
Sort by score descending to see your strongest pages; sort by score ascending to find your worst-performing pages among those you track.
Go deeper: Visibility > Citation Sources + Analytics > Crawler Analytics
Cross-reference with Visibility > Citation Sources to see the citation breakdown per URL.
The Citations view shows you which AI providers are citing a specific URL, how many times, and in response to which types of queries. A page with a high composite score but all citations concentrated on one provider is more fragile than a page with citations spread across three providers — diversification matters for score stability.
Then check Analytics > Crawler Analytics for the same URL to verify the crawler visit component: a page that is frequently crawled but rarely cited is a content quality problem. A page that is rarely crawled despite having citations suggests the crawl frequency component is dragging down an otherwise strong score.
Complete the picture: Strategy > GSC
Connect Strategy > Google Search Console to see the GSC authority component of the score.
A page that ranks in the top 10 in traditional search for its primary query also tends to have a strong AI performance score — AI models frequently use traditional search authority as a quality proxy. Pages in positions 1–3 for high-intent queries are disproportionately likely to be cited by AI. The GSC view lets you identify pages that are strong in traditional search but weak in AI citations — these have the infrastructure for a strong AI score but need targeted citation-building to close the gap.
What to Look For
Page Performance — Score Composition Breakdown
The AI performance score is designed to be actionable, not just descriptive. Each component points to a specific improvement lever: low citation rate suggests content quality or source authority improvements; low crawler frequency suggests crawl discoverability fixes (sitemap, internal linking, robots.txt); low referrer traffic from AI tools suggests citation link presence (mentioned but not linked); low GSC ranking suggests broader authority building is needed first.| Element | What it tells you |
|---|---|
| Overall AI performance score | Composite health of this URL across all AI performance dimensions |
| Citation rate component | How often AI models choose this URL as a trusted source |
| Crawler visit component | How regularly AI bots are indexing and re-evaluating this page |
| Referrer traffic component | Whether AI tool users are actually clicking through from AI-cited links to this page |
| GSC authority component | Traditional search rank as a proxy for domain authority on this topic |
Citations — Provider and Query Distribution
A strong AI performance score should show citation diversity: the page is cited by at least 2–3 different AI providers, in response to a variety of query types (not just one narrow question). Single-provider citations are vulnerable to that provider changing its sourcing behavior — distributed citations are more stable and indicate genuine authority.Pro Tip: Sort Page Performance by the citation rate component (not the overall score) to find pages that are trusted by AI models but not yet driving traffic. These pages need their citation links to be formatted as hyperlinks, not just brand mentions — this is often a quick technical fix that converts latent AI citations into actual referrer traffic.
Filters That Help
| Filter | How to use it for this question |
|---|---|
| URL | Drill into a specific page to see all four score components and the time-series trend |
| Provider | Check how the score components vary by AI provider — a page strong on Perplexity but weak on ChatGPT may need content adjustments specific to ChatGPT’s preferred source types |
| Period | Track score evolution over time to measure the impact of content improvements |
How to Interpret the Results
Good result
Your 10 most strategically important pages all have AI performance scores above 40. Citation rate is present on at least 2 AI providers per page. Crawler visit frequency shows weekly or more frequent re-visits from at least 2 major AI bots. Referrer traffic from AI tools is measurable (not zero) for pages with high citation rates. The GSC authority component is above 50 for pages in competitive topics, reflecting strong traditional search ranking as a foundation.Needs attention
Core product and landing pages have AI performance scores below 15. Citation rate is zero despite crawler visits — confirmed content quality issue. Referrer traffic is zero despite recorded citations — confirmed citation link format issue (brand name mentioned but no hyperlink). GSC authority component is low across the board — indicates authority-building must precede citation-building, as AI models will not consistently cite pages that traditional search also ranks poorly.Example
Scenario: An e-commerce marketplace selling electronics and home appliances wants to understand which of their 65 tracked product category pages have the strongest AI performance and which need urgent attention.
- Open Analytics > Page Performance and sort by score descending. The top 3 pages are: a buying guide (“Best noise-cancelling headphones 2025” — score 71), a category comparison (“Robot vacuums under 500 euros compared” — score 63), and the marketplace’s “deals and promotions” landing page (score 48). All three make sense — these are the pages where shoppers make purchasing decisions, and AI models cite them frequently in shopping-intent queries.
- Sort by score ascending. The bottom 5 include several high-traffic product category pages (scores 3–8) that drive significant revenue. Open the “smart home devices” category page with a score of 5. The breakdown shows: citation rate near zero, but crawler visit frequency is healthy (GPTBot visits weekly, PerplexityBot visits every 2 days). This is a content quality problem — the crawlers find the page but AI models are not using it as a source because the page is primarily a product grid with no editorial content.
- Check Visibility > Citation Sources filtered to the smart home devices page. The page is never cited. Open Strategy > GSC for the same URL — it ranks position 32 for “best smart home devices.” The GSC data confirms: the page lacks the editorial depth and traditional search authority that AI models use as a quality signal. The action: add a curated buying guide section with comparison tables and expert recommendations above the product grid, then build internal links from the existing high-performing buying guides to boost authority.
Go Further
Per-page score dashboard
Build a per-page AI score dashboard in Looker Studio using the source-urls data source
How AI scores are calculated
Read the Page Performance documentation to understand how unified AI scores are calculated for each page
Pull per-page scores via API
Use the source-urls endpoint to programmatically pull unified AI scores for every page on your site
Related Questions
Which URLs does my brand get cited from most?
See your top-cited URLs and what query types are driving those citations
Which pages get the most traffic from AI tools?
Combine AI performance scores with actual referrer traffic data per page
How quickly do AI models pick up new content?
Track the ramp-up curve for new pages from publication to full AI score