The Evolution of Search Metrics: Bing’s Move Into AI Attribution
The landscape of search engine optimization is undergoing its most significant transformation since the advent of mobile-first indexing. As artificial intelligence becomes deeply integrated into how users find information, the traditional metrics of clicks, impressions, and rankings are no longer the only markers of success. In a major move to provide transparency in this new era, Microsoft has officially launched the AI Performance report in Bing Webmaster Tools.
Currently in public beta, this new dashboard is designed to help webmasters, SEO professionals, and content creators understand how their work is being utilized by generative AI. Specifically, it tracks how often a website’s content is cited as a source within Microsoft Copilot, Bing’s AI-powered summaries, and various third-party partner integrations that leverage Bing’s index to ground their AI models. This launch marks a pivotal moment for “Generative Engine Optimization” (GEO), providing the first real set of data points for those trying to optimize for the AI-first web.
What is the AI Performance Report?
The AI Performance report is a dedicated dashboard located within the Bing Webmaster Tools suite. Its primary function is to track “citations.” In the world of generative AI, a citation is a link or a reference that the AI provides to indicate where it retrieved the information used to generate its response. When a user asks Microsoft Copilot a question, the AI scans the web to find reliable data. If it uses your website to formulate that answer, the AI Performance report will log that event.
For years, SEOs have relied on the Performance report in Bing Webmaster Tools (or Google Search Console) to see which keywords drove traffic. The AI Performance report operates on a different logic. It doesn’t necessarily track “search queries” in the traditional sense; instead, it tracks “grounding queries”—the prompts or searches that led the AI to use your specific pages as the factual foundation for its output.
Microsoft first began testing this feature in late January, and its full release into public preview signifies the company’s commitment to an open ecosystem. By showing publishers how they contribute to the AI’s knowledge base, Microsoft is attempting to bridge the gap between AI consumption and content creation.
Key Metrics Explained: Decoding the Dashboard
The new dashboard introduces several specific metrics that differ from traditional search analytics. To gain value from the AI Performance report, it is essential to understand what each of these data points represents and how they interact with one another.
Total Citations
This is the headline figure of the report. It represents the total number of times any page from your website was cited as a source in an AI-generated answer during a specific period. It is the AI equivalent of an “impression,” but with a higher level of significance, as it implies your content was deemed authoritative enough to serve as a primary source for the AI’s response.
Average Cited Pages
This metric calculates the daily average of unique URLs from your site that are referenced across AI experiences. If you have a large content hub, this number helps you understand the “breadth” of your authority. A high number of total citations coming from only one or two pages suggests you have a few “blockbuster” articles, whereas a high average of cited pages indicates that the AI views your entire domain as a reliable resource across multiple topics.
Grounding Queries
Perhaps the most valuable part of the report for SEOs, Grounding Queries are the specific phrases or questions users typed into Copilot or Bing that triggered the AI to use your content. This functions similarly to keyword data but offers a glimpse into the conversational nature of AI interactions. By analyzing these queries, publishers can see the exact intent their content is satisfying in the eyes of the AI.
Page-Level Citation Activity
This section breaks down performance by individual URL. It allows you to see which specific pages are the workhorses of your AI visibility. If a page is getting high citations but low traditional search traffic, it may be because it is highly factual and well-structured—ideal for AI grounding—even if it isn’t ranking in the top three of a standard SERP.
Visibility Trends Over Time
Like any performance tracker, the AI Performance report includes a timeline view. This allows webmasters to see if their AI visibility is growing or shrinking. It is particularly useful for tracking the impact of content updates or seeing how changes in the AI models (like an update to Copilot) affect how often your site is referenced.
The Rise of Generative Engine Optimization (GEO)
With the release of this tool, Microsoft is essentially legitimizing the field of Generative Engine Optimization. For a long time, SEO was about matching keywords and building backlinks to climb a list of ten blue links. GEO is different; it is about ensuring your content is the most “extractable” and “verifiable” source for an LLM (Large Language Model).
Microsoft has explicitly stated that this tool is an early step toward helping publishers navigate this shift. To perform well in AI citations, the requirements are slightly different than traditional SEO. While standard SEO best practices still apply, GEO places a heavy emphasis on:
- Information Density: Providing direct answers to complex questions.
- Structural Clarity: Using H2 and H3 tags, bullet points, and tables that AI can easily parse.
- Factual Accuracy: AI models are increasingly tuned to prefer “grounded” and “verified” facts over fluff.
- Entity Representation: Ensuring that the people, places, and products mentioned on your site are clearly defined so the AI can connect them to its existing knowledge graph.
The Missing Piece: The Traffic and Click-Through Dilemma
While the AI Performance report is a welcome addition to the webmaster’s toolkit, it is not without its limitations. The primary criticism from the SEO community is the lack of click-through data. Currently, the report shows you that you were cited, but it does not tell you if the user actually clicked the citation to visit your website.
In traditional search, an impression is only valuable if it leads to a click. In the AI world, if Copilot answers the user’s question entirely within the chat interface using your data, the user may have no reason to visit your site. This “zero-click” reality is the biggest fear for modern publishers. Without click data, it is difficult to calculate the ROI of AI visibility. Is a citation worth as much as a search result? Or is it a form of “digital sharecropping” where the AI gets the utility and the publisher gets a vanity metric?
Microsoft has acknowledged these concerns, stating that they plan to continue evolving these capabilities. The hope is that future iterations of the report will include referral traffic data specific to AI citations, allowing publishers to see exactly how much traffic Copilot is actually sending their way.
Practical Strategies: How to Use the AI Performance Data
Despite the lack of click data, the AI Performance report provides actionable insights that can be used to refine a content strategy. Here are several ways to put this data to work immediately:
1. Identify Your “AI Stars”
By looking at the page-level citation activity, you can identify which of your existing articles are most “AI-friendly.” Analyze these pages to see what they have in common. Are they long-form? Do they use a specific type of schema? Do they have a “Key Takeaways” section? Once you find the formula that works for the AI, you can replicate it across other sections of your site.
2. Content Gap Analysis via Grounding Queries
Look at the grounding queries that are triggering citations. If you see queries that are related to your niche but for which you are *not* being cited, it indicates a content gap. You can create new content specifically designed to answer those prompts. Conversely, if you are being cited for queries that don’t lead to your most profitable pages, you might need to find ways to interlink those cited pages with your conversion-focused content.
3. Optimizing Low-Performing Pages
If you have high-quality, long-form content that is not appearing in the AI Performance report, it may be a structural issue. Use Microsoft’s guidance to improve the clarity and completeness of these pages. This includes using evidence-backed claims and ensuring your headers are descriptive rather than creative. AI prefers “How to bake a sourdough loaf” over “The magic of rising yeast.”
4. Monitoring Model Sensitivity
AI models are updated frequently. By monitoring the “Visibility Trends,” you can see if an update to Copilot has caused your citations to drop. If your traditional rankings remain stable but your AI citations fall, it suggests that the AI’s selection criteria have changed, and you may need to adjust your content structure to regain that visibility.
Bing vs. Google: The Transparency War
It is worth noting the competitive landscape in which this tool was released. Google has been testing its own AI-integrated search results (formerly SGE, now AI Overviews) for quite some time. However, Google has been much more guarded with its data. While Google Search Console has integrated some AI-related data into its performance reports, it hasn’t yet provided a dedicated, granular “AI Performance” dashboard like Bing has.
By launching this tool in beta, Microsoft is positioning Bing Webmaster Tools as the more transparent and publisher-friendly option. This is a strategic move to court the SEO community and encourage them to spend more time optimizing for Bing. For a search engine that has long sat in the shadow of Google, providing “first-of-its-kind” data is a powerful way to gain mindshare among digital marketers.
Conclusion: Preparing for the Post-Search World
The launch of the AI Performance report in Bing Webmaster Tools is more than just a new feature; it is a signal of where the industry is heading. We are moving away from a world of “search” and into a world of “discovery” and “synthesis.” In this new environment, your value as a publisher is determined by how well your content can ground an AI’s response.
While the tool is still in its early stages and lacks the crucial click-through metrics that many demand, it provides a much-needed window into the “black box” of generative AI. By understanding which pages are cited and which queries trigger those citations, SEOs can begin to build a framework for the future of the web.
As Microsoft continues to refine this tool, the focus will likely shift toward attribution and conversion. For now, the best course of action for any webmaster is to dive into the data, identify what the AI likes about their content, and double down on clarity, structure, and authority. The era of Generative Engine Optimization is officially here, and the AI Performance report is the first roadmap we have to navigate it.