For more than two decades, search visibility meant one thing to most businesses: where do you rank on the results page? Tools from the search engines themselves, such as Google Search Console and Bing Webmaster Tools, have long provided this information, helping marketing teams build strategies around impressions, clicks, and keyword positions.
But a quiet shift has been underway. Increasingly, people are obtaining their answers directly from AI. They ask a chatbot, read a generated summary, or follow a Copilot response. In many cases, they never visit a traditional search results page at all. For publishers, the underlying need remains unchanged. They still need to know how visible their content is. What has changed is where that visibility is measured. It is no longer enough to know whether people can find your website in search results. You also need to know whether AI systems are using your content to answer questions in your field.
Microsoft has now made that second kind of visibility measurable. AI Performance, a new feature inside Bing Webmaster Tools, is now live and gives website owners a clear picture of how often their content is cited across Microsoft Copilot. For the first time, businesses can see not just that their pages were crawled, but that they were referenced as a source inside an AI-generated answer.
This is a practical early step toward what the industry is calling Generative Engine Optimisation, or GEO. At present, GEO is heavily biased toward visibility: making your content more likely to be discovered, trusted, and cited by generative AI systems. It is not yet about every capability AI can offer, such as direct transactions, agentic actions, or deep reasoning tasks. It is about being seen and referenced in the answers AI produces. And while that focus is narrow, it is no small thing. The arrival of built-in reporting inside a major search platform means GEO is no longer a theoretical idea. It is something your team can act on today.
Why AI Citations Matter for Business
When a customer searches for a product category, a service area, or a problem your business solves, the ideal outcome has always been the same: your brand appears in a position of trust. In the era of blue links, that meant ranking near the top. In the era of AI answers, it increasingly means being the source that an AI system names when it generates a response.
Being cited in an AI answer carries several forms of value. It signals authority. It keeps your brand present during the research phase of a buying journey. It can drive traffic when the answer includes a link. And even when no click happens, the citation builds familiarity, which influences decisions later.
The challenge until now has been opacity. Businesses could see search impressions and clicks in Bing Webmaster Tools, but they had no equivalent report for AI. They could not tell whether Copilot had referenced their help article, whether Bing’s AI summary had pulled from their product page, or whether their thought leadership was being used to answer industry questions. They also had little idea how AI systems understood their website, or which user questions their content was being used to answer. That gap made it hard to invest confidently in GEO.
Of course, many organisations have treated SEO as a proxy for GEO, and that is not wrong. Strong search performance and well-structured content often help with AI citations too. But it is still a proxy. Until now, teams have had to assume that what works for search will also work for AI answers. Microsoft’s AI Performance report removes that blind spot.
What the AI Performance Dashboard Shows
For webmasters already familiar with tools such as Google Search Console and Bing Webmaster Tools, the AI Performance dashboard will feel like a natural extension rather than a foreign interface. The layout, date range selectors, and metric cards follow familiar patterns. The difference is simply that the data now covers AI-generated answers instead of only traditional search results.
The dashboard gives a consolidated view of how your site participates in AI-generated answers across supported Microsoft experiences.
Here are the main metrics it provides.
Total Citations: This shows the total number of times your content was displayed as a source in AI-generated answers during the date range you select. It is a count of references, not a measure of how prominent each reference was. That distinction is important. A high number means your content is being used frequently. It does not necessarily mean it appeared at the top of every answer.
Average Cited Pages: This shows the average number of unique pages from your site that were cited per day over the selected period. Because the data is aggregated across multiple AI surfaces, this reflects overall citation patterns rather than the role of any individual page in a single answer.
Grounding Queries: These are the key phrases that the AI used when retrieving content that later appeared as citations. They offer a window into the questions and topics for which your site is considered relevant. Microsoft notes that this is a sample of activity rather than a complete record, and that the metric will be refined as more data is processed.
Intents: Grounding queries are classified into broader intent categories such as Informational, Commercial, Navigational, Learn and Solve, Research, Creation, and Local. This helps publishers understand not just which queries triggered citations, but the kind of user context behind them. An e-commerce publisher might see strong visibility in comparison-oriented interactions, while an educational publisher might find their content appearing in research or learning contexts.
Topics: Related grounding queries are grouped into thematic clusters, so publishers can see which subject areas are driving citation activity rather than analysing isolated keywords. Queries such as “solar panels,” “solar energy efficiency,” and “residential solar installation” might all roll up under a broader topic such as Solar Energy. This makes it easier to align editorial planning with the way AI systems semantically organise information.
Citation Share: This shows your site’s percentage of all citations shown for a specific grounding query. If your site received three out of ten citations for a query, your Citation Share would be 30%. It is an observational metric only. It is not a ranking system, a competitive scoreboard, or a traffic share figure. It does not reveal competitor domains or assign quality scores. It simply helps you understand how much of the citation space your content occupies for a given query.
Visibility Trends Over Time: A timeline view lets you see whether your AI citations are growing, flat, or declining. That trend line is valuable for connecting content efforts with outcomes. A new content push, a site relaunch, or a technical fix may show up here weeks or months later.
Compare: This lets you overlay a previous time period onto the current view, so you can observe how citation activity, intent patterns, topic coverage, or Citation Share change over time. You might compare the current 30 days against the prior 30, or choose custom ranges that align with content campaigns or site changes.
Page-level Citation Activity: This report shows which specific URLs on your site are cited most often. It helps you identify your high-performing pages in the AI context, compare them with the pages you consider strategically important, and spot content that might deserve more attention.
Microsoft also makes clear that Bing respects robots.txt and other content owner controls. If you do not want your content included, the existing mechanisms still apply.
What This Means for Content Strategy
The arrival of AI Performance reporting changes how marketing and publishing teams should think about their content. Traditional SEO remains essential, but it is no longer the whole game. Businesses now need to optimise for two discovery layers: search results and AI-generated answers.
These layers are related but not identical. A page can rank well yet rarely be cited by AI. Another page might rank modestly but be quoted often because it is structured in a way that AI systems find easy to use. AI Performance makes it possible to tell the difference.
This creates an opportunity to be more strategic. Instead of guessing what content AI prefers, you can study your own citation data. You can see which topics trigger references, which formats are most often used, and where your coverage might be thin. Unlike a simple search query report, AI Performance also shows which pages were actually cited in the AI answer. That gives you a clearer context about what customers are looking for and how your website satisfies those needs.
The Intents and Topics layers make this even more powerful. Intent categories reveal what customers are trying to accomplish when AI cites your content. Are they learning, comparing, buying, navigating, or solving a local problem? That tells you how your website is being used in real customer journeys. Topic clusters, meanwhile, show how AI systems position your brand within broader subject areas. Instead of seeing your site as a collection of isolated keywords, you begin to see it as a participant in the conversations your customers care about. A publisher focused on energy might discover that AI treats them as a trusted voice across solar power, home batteries, and grid policy. A software company might find that its strongest AI visibility sits in implementation and migration topics rather than product comparison. These patterns reshape how you plan content, prioritise editorial investment, and define your competitive niche.
You can then make content decisions based on evidence rather than intuition.
What AI Performance Does Not Yet Cover
As useful as AI Performance is, it is important to understand its boundaries. Otherwise, publishers may mistake a partial view for the complete picture.
Bing is not the only AI chatbot in the market. Microsoft Copilot and Bing AI summaries are significant experiences, but they are not the whole AI landscape. Other major AI assistants, search products, and chatbot platforms are not currently offering the same first-party citation reporting to publishers. A business that receives most of its traffic from other AI ecosystems will not see that activity reflected here. AI Performance is, therefore, a window into Microsoft’s corner of the AI web, not a universal dashboard for all generative AI citations.
Local and open-weight models are invisible. The dashboard only covers online chatbot services that Microsoft operates or partners with. It does not capture what happens on locally run open-weight large language models. A user running an LLM on their own machine, a private corporate deployment, or an edge device may still cite your content, but those citations and queries will never be reported. That slice of AI usage is growing, and it is inherently opaque to publishers.
These limitations do not make the tool useless. They simply mean AI Performance should be treated as one signal among many. It tells you something important about one set of AI surfaces. It does not tell you everything about how AI systems are using your content everywhere.
The Bigger Picture: Transparency Between AI and the Open Web
Microsoft describes AI Performance as an early step toward greater transparency between AI systems and the open web. That framing is worth taking seriously. For years, publishers have worried that generative AI would use their content without credit, attribution, or visibility. Features like AI Performance do not solve every concern, but they do begin to close the information gap.
By showing publishers when and where their content is cited, Microsoft is creating a foundation for a healthier relationship between AI systems and content creators. Publishers can see value. They can optimise. They can decide whether they want to participate. And the webmaster community can give feedback that shapes how these tools evolve.
This also sets a competitive standard. As more businesses begin measuring AI citations, the brands that invest early in structured, authoritative, up-to-date content will build a measurable edge. Those who ignore GEO risk become invisible in the places where an increasing share of customer research happens.
What Leaders Should Do Now
The launch of AI Performance is not a reason to abandon traditional SEO. Search results still matter, and they will continue to drive significant traffic for years. But it is a reason to broaden your view of visibility.
Here is a practical action plan for leadership and marketing teams.
First, verify your site in Bing Webmaster Tools. If you have not already done so, claim your property and explore the new AI Performance report. Even if your citations are low at first, having baseline data is essential.
Second, compare cited pages with business priorities. Are the pages being cited the same ones that matter most to your strategy? If your flagship product page is rarely cited while a secondary blog post dominates, you have a strategic gap to address.
Third, study grounding queries for content opportunities. The phrases that trigger AI citations reveal what your audience is actually asking. Use them to inform editorial calendars, FAQ expansions, and new product or service content.
Fourth, audit your most cited pages for freshness and structure. Make sure they are current, well-organised, and supported by evidence. These pages are already winning in the AI context, so small improvements can have an outsized impact.
Fifth, consider IndexNow for faster updates. If your content management system or hosting platform supports it, enable IndexNow so that changes are discovered quickly across search and AI experiences.
Sixth, add Microsoft Clarity if you have not already. It is free, privacy-focused, and now integrates AI Visibility data that complements Bing Webmaster Tools. The combination gives you both discovery and behavioural insight.
Seventh, treat GEO as a shared responsibility. It should sit across content, SEO, product marketing, and web operations. No single team owns it, and the businesses that coordinate well across functions will move fastest.
A New Metric for a New Era
AI Performance in Bing Webmaster Tools is more than a feature update. It is the arrival of a new metric for a new era of discovery. Just as impressions and clicks once helped businesses understand their place in search, citations and grounding queries will help them understand their place in AI.
The shift is still early. The dashboard will evolve, the definition of a citation will be refined, and best practices will mature. But the direction is clear. Businesses that create structured, trustworthy, current content will be the ones most often cited by AI systems. And those citations will become a new form of currency in the competition for customer attention.
If you lead a marketing, publishing, or content team, now is the time to add AI citations to your reporting rhythm.
Leave a Reply