Microsoft has announced an exciting update to its AI Performance dashboard within Bing Webmaster Tools. This new feature allows publishers to directly associate grounding queries with the specific web pages referenced in AI-generated responses. For the first time, content creators can pinpoint the exact material that enhances their visibility within AI search systems.
On March 23, 2026, Krishna Madhavan, Principal Product Manager at Microsoft AI and Bing, shared this update, which comes just a month after the initial public preview of the AI Performance dashboard on February 10, 2026. The new Grounding Query – Pages Mapping feature is now being launched worldwide, marking it as the most requested addition since the dashboard debuted.
“The AI Performance report introduced in Bing Webmaster Tools has proven to be a significant success,” Madhavan noted in a LinkedIn post. “One of the most anticipated features, ‘Grounding Query – Pages Mapping,’ is now being rolled out globally.”
The timing of this rollout is crucial. While the original dashboard provided publishers with aggregate citation counts and a list of grounding queries in separate views, this new mapping capability integrates the two features. Publishers can now select any grounding query to instantly view which pages of theirs were referenced in response to that query. Alternatively, clicking on a specific URL will show all grounding queries that led to citations for that page. This two-way connection is a significant innovation – a single grounding query can refer to multiple pages, and a single page can be connected to multiple grounding queries at the same time.
Understanding Grounding and Its Importance
To grasp the importance of this update, it’s essential to know what grounding entails in the realm of AI search. Grounding involves AI systems retrieving up-to-date, authoritative web content before generating answers, rather than depending solely on training data. It acts as the bridge linking user inquiries to relevant information available on the internet.
According to Microsoft, their grounding technology supports almost all major AI assistants available. This technological infrastructure is responsible for real-time data retrieval, verification of sources, and enabling AI systems to cite specific pages when formulating responses. The initial announcement of the AI Performance dashboard in February was Microsoft’s first structured approach to revealing this process to publishers, alongside an in-depth explanation of grounding as foundational infrastructure by Jordi Ribas, Corporate Vice President of Search and AI.
The February launch introduced four key metrics within Bing Webmaster Tools: total citation counts, average cited pages per day, grounding queries, and page-level citation activity. These metrics revealed surprising trends for some publishers, showing that a webpage could receive numerous citations in AI responses without generating a single click, while pages that drew consistent organic traffic might not appear as citation sources at all. This discrepancy is not a fault; it highlights how AI-generated answers can often fulfill user information needs directly, eliminating the necessity for clicks on links.
This divergence creates an analytical gap. Since citations and traffic now function as distinct performance metrics, the tools needed for optimization must also differ. The Grounding Query – Pages Mapping feature begins to bridge that gap by offering detailed, reciprocal data to inform editorial decisions based on AI citations rather than traffic figures alone.
Practical Applications of the Mapping Feature
Within the AI Performance report in Bing Webmaster Tools, users now see two interconnected views: Grounding Queries and Pages. Previously, these appeared as two separate tables. With the new capability, selecting any entry in either view will automatically filter and cross-reference the data in the other.
For instance, if a publisher notices the grounding query “sustainable packaging regulations Europe 2025” surfacing frequently, they can click on it to see precisely which URLs Bing cited in generating answers for related questions. If those citations point to a singular outdated article, the publisher has a strong, data-driven reason to update that content. Conversely, if the citations are distributed across multiple pages with similar content, it indicates an opportunity for consolidation.
The reverse direction is equally insightful. A content team might be surprised to discover that a product comparison page—despite generating minimal organic traffic—is being cited frequently in AI responses. Without the mapping feature, that vital signal might have been hidden in aggregate data. Now, the specific queries driving citations become clear, offering insight into how AI systems gauge the page’s relevance.
Madhavan stated that this feature was implemented in direct response to the feedback received after the February launch. The speed of its rollout—from launch to global availability in about five weeks—reflects the level of engagement from the publisher community with the AI Performance tools.
Market Context: $750 Billion and Half of Consumers
Microsoft emphasizes the business implications of AI search visibility with a significant statistic. Citing McKinsey research, they report that half of consumers are already utilizing AI-powered search, a trend poised to influence $750 billion in revenue by 2028. Regardless of whether these projections fully materialize, the trend described is evident: AI-generated responses are increasingly intercepting queries that once led users directly to brand websites.
This shift introduces a new category of brand risk. A company that excels in traditional search but whose content is not cited in AI responses may find its visibility diminishing as AI-assisted search picks up steam. On the flip side, brands whose content is frequently referenced in AI answers could enhance their credibility with users earlier in the decision-making process, even without direct traffic.
The original AI Performance dashboard noted that Microsoft’s advertising business surpassed $20 billion in annual revenue as AI-powered search features gained traction, underscoring the financial significance linked to AI search visibility.
Key Insights for Publishers Regarding the Data
Microsoft has clearly outlined one important limitation: the metrics in the AI Performance dashboard provide samples of overall citation activity, rather than exhaustive records. The grounding queries presented are representative samples, not an all-inclusive log of every query that triggered a citation. Publishers should consider this sampling limitation when making any editorial decisions based on the data.
Furthermore, Bing adheres to all content ownership preferences expressed through robots.txt and other supported control mechanisms. Content excluded from AI snippets via robots.txt or the data-nosnippet attribute will not appear in citations and, therefore, will not show up in the AI Performance dashboard.
The dashboard is part of a broader suite of tools Microsoft has been gradually enhancing over the past 18 months. Bing Webmaster Tools extended its historical search performance data from 6 months to 16 months in October 2024, and then to 24 months in August 2025. Device and country filtering features were incorporated in August 2025, and sitemaps were identified as critical resources for AI search discoverability in a detailed technical announcement during that month. The AI Performance dashboard and the new Grounding Query – Pages Mapping feature represent the next advances in this platform, transitioning from crawl-and-index signals to data focused on AI-specific optimization.
The LinkedIn post announcing the mapping feature garnered significant engagement from professionals in search and content marketing. For instance, Cem Ozcelik, Growth Manager at Adsby and Visby, inquired whether this data would eventually be available through the Bing Webmaster Tools API, indicating that practitioners are already considering ways to integrate citation data into automated reporting and monitoring frameworks. Madhavan did not confirm API availability but acknowledged the inquiry, with Ozcelik expressing hope that it would be included in the API roadmap.
Richard Nazarewicz, Head of Search Strategy and SEO Product at BBC Global News, asked if the data could be further filtered by the origin of traffic, particularly distinguishing between IndexNow and Bingbot. This query highlights a technically detailed distinction: IndexNow is a protocol allowing publishers to notify search engines of content updates in real time, while Bingbot refers to Bing’s crawling infrastructure. Whether these two pathways lead to differing citation rates remains an open question that the current dashboard does not address.
Eduard Blacquière, a freelance SEO consultant and co-founder of SEO Pro Academy, simply stated, “Love it, also how quickly you are shipping this!”
The volume of technical follow-up inquiries indicates a general unfamiliarity with the underlying data model, even among seasoned search professionals. Grounding query data does not have a direct counterpart in traditional search analytics. There is no “grounding query” column in Google Search Console or any established SEO tools. Microsoft’s release essentially introduces a new category of performance data that practitioners are only beginning to explore.
The Future of AI Performance Tools
The announcement regarding the mapping feature indicates that the AI Performance dashboard is still evolving. Madhavan and Meenaz Merchant, Partner Group Product Manager at Microsoft Bing, suggest that Microsoft views this tool as part of a broader Generative Engine Optimization toolset that will continue to develop. Publishers are directed to the Microsoft Advertising AI Web hub for further guides and optimization resources.
Specific details about upcoming metrics or features were not disclosed. Currently, the dashboard includes total citations, average cited pages, grounding queries, page-level citation activity, visibility trends over time, and the new grounding query-to-page mappings. Potential future directions—though not confirmed by Microsoft—could include data on citation positioning within AI-generated answers, benchmarking citation rates against competitors, or further integration with IndexNow signals.
This trajectory suggests that Microsoft is treating AI search visibility as a separate discipline from traditional SEO, necessitating its own specialized toolset. The Copilot Search launched in Bing in April 2025, merging conventional search with generative AI capabilities. Similarly, Bing Webmaster Tools incorporated Copilot-powered assistance in a limited preview in October 2024. Each component reinforces the others, setting the stage for a comprehensive ecosystem.
For marketing professionals, the implications are clear. AI citation visibility and traditional search visibility are now distinct signals requiring separate monitoring. With the global rollout of the Grounding Query – Pages Mapping feature, publishers can systematically track and optimize both dimensions of their search presence.
Key Milestones
- January 31, 2025 – Microsoft integrates Grounding with Bing Search into Azure AI Agent Service, enhancing real-time web data access for AI applications.
- April 4, 2025 – Microsoft launches Copilot Search in Bing, blending traditional search with generative AI.
- August 7-8, 2025 – Microsoft adds device and country filtering to Bing Webmaster Tools and extends data retention to 24 months.
- August 9, 2025 – Microsoft repositions sitemaps as essential infrastructure for AI-powered search.
- October 15, 2025 – Microsoft introduces the data-nosnippet HTML attribute for Bing, granting publishers control over AI snippets.
- October 20, 2024 – Bing Webmaster Tools extends search performance data from 6 to 16 months and launches Recommendations feature.
- February 10, 2026 – Microsoft launches AI Performance dashboard in public preview, the first tool providing publishers visibility on citations in AI-generated answers.
- February 11, 2026 – Microsoft publishes updated AI marketer’s guide, elucidating how AI search surfaces brands.
- February 12, 2026 – Microsoft explains grounding as the foundational infrastructure that powers nearly all major AI assistants.
- March 23, 2026 – Microsoft announces the global rollout of Grounding Query – Pages Mapping within the AI Performance dashboard in Bing Webmaster Tools.
Summary
Who: Microsoft, represented by Krishna Madhavan (Principal Product Manager, Microsoft AI and Bing) and Meenaz Merchant (Partner Group Product Manager, Microsoft Bing), announced the update aimed at advertisers, publishers, SEO professionals, and content marketers utilizing Bing Webmaster Tools.
What: The Grounding Query – Pages Mapping feature is now available globally in the AI Performance dashboard within Bing Webmaster Tools. This new capability allows bidirectional navigation between grounding queries and the specific pages cited in AI-generated responses, enabling users to determine which content drives AI citation visibility and which queries reference specific pages.
When: This feature was announced and began its global rollout on March 23, 2026, approximately one month after the AI Performance dashboard itself was launched in public preview on February 10, 2026.
Where: Available within Bing Webmaster Tools, accessible to any publisher with a verified site on the platform. The citations tracked include Microsoft Copilot, AI-generated answers in Bing, and select partner integrations utilizing Microsoft’s grounding technology for AI responses.
Why: The feature addresses direct user demand, being the most-requested enhancement following the February launch. Additionally, it addresses a structural gap in publisher analytics: as AI-generated responses intercept queries that previously directed users to brand websites, citation visibility emerges as a distinct and measurable performance signal that requires dedicated data tools. McKinsey research cited in the announcement indicates that AI-powered search already impacts half of consumers and could influence $750 billion in revenue by 2028.
Share this article
