Bing Webmaster Tools AI Performance, the complete guide to tracking your AI citations in 2026

Until now, measuring visibility in AI-generated responses was a kludge, like querying ChatGPT or Copilot manually, using third-party tools like Profound, or simply crossing your fingers. With AI Performance, Bing crosses a threshold by offering the first structured, official report dedicated to AI citations.

It is not trivial. Google folds its AI Overviews data into the global Performance report in Search Console, with no isolation and no per-URL count. Bing takes the lead by isolating these metrics in a dedicated space, an approach the entire SEO and GEO industry was waiting for.

For search professionals, this is the confirmation that Generative Engine Optimization (GEO) is no longer speculation. It is a measurable discipline, with its own KPIs and tracking tools. As Microsoft's official announcement summarizes, AI Performance represents a first step toward native GEO tools inside webmaster platforms .

Once your site is verified in Bing Webmaster Tools, the AI Performance report is reachable from the navigation menu on the left. By default, it shows data for the past 30 days. Here is an overview of the dashboard.

The dashboard is built around 5 key metrics, shown below.

Grounded queries are probably the most interesting metric in the report, and the most misunderstood. Unlike classic search queries, they do not represent what the user typed.

When a user asks Copilot a question, the system internally rephrases that question into one or more micro-queries optimized for content retrieval. Those rephrasings are what Bing calls "grounded queries". For example, a user asking "how to improve my marketing team's productivity" could generate grounded queries like "marketing team productivity strategies 2026" or "marketing project management tools".

According to Otterly.ai's data, a mid-sized site can generate several hundred unique grounded queries over a 3-month period. Their analysis surfaced 647 unique queries triggering 30,398 grounding events across 173 pages .

On June 16, 2026, Microsoft expanded the AI Performance report with four new dimensions, rolled out in preview globally. They turn a raw list of grounded queries into a genuine strategic lens. Here is what each one brings.

Microsoft notes that the Intents and Topics classifications are still maturing and will sharpen as data accumulates. The GEO recommendations previewed at SEO Week (crawlability, structured data, indexing), however, have not shipped yet, nor have click data and API access, both still expected during 2026 .

The question every SEO is asking is how Bing AI Performance compares to what Google offers in Search Console. Here is a detailed comparison.

The takeaway is clear, Bing has taken the lead on AI-specific citation tracking. That does not mean Google will not catch up, but in the meantime, Bing Webmaster Tools is the reference tool.

Here are the steps to reach the dashboard and start analyzing your AI citations:

• Go to bing.com/webmasters and sign in with your Microsoft account.
• If not done yet, verify ownership of your site through XML file, meta tag, or DNS record.
• Once verified, click "AI Performance" in the navigation menu on the left.
• The dashboard shows the past 30 days by default. Adjust the date range to your needs.
• Export grounded query data and per-page citations to analyze them in a spreadsheet or BI tool.

Having access to the data is not enough. Here are 7 concrete strategies to grow your presence in Bing Copilot AI responses, based on GEO best practices and the signals Microsoft favors:

• Structure for extraction by opening each article with a 40-80 word answer capsule that directly answers the question. AI systems favor content whose blocks are easy to extract. Phrase your H2s as real questions that mirror search intents.
• Add quantitative data and expert quotes, because pages enriched with verifiable statistics and attributed citations earn significantly more references. The SE Ranking study shows that pages with 19+ data points reach 5.4 citations on average.
• Implement structured markup, since Article, FAQ, HowTo, and Product schemas help RAG (Retrieval-Augmented Generation) systems understand the contextual relationships in your content. Schema.org markup is especially critical for Copilot citations.
• Keep content fresh with IndexNow, which lets you instantly notify search engines (including Bing) when content is added, modified, or removed. This ensures AI cites the most recent version of your pages. Pages updated in the past 3 months earn 6 citations versus 3.6 for stale content.
• Allow AI crawlers in your robots.txt by checking that Bingbot and OAI-SearchBot (and ChatGPT-User) are not blocked. Many sites accidentally block these crawlers while allowing Googlebot, cutting themselves off from AI visibility.
• Build topical authority, because AI platforms favor content from recognized sources. Diversify your presence on trusted platforms (Reddit, Quora, G2, Trustpilot), earn quality backlinks, and aim for mentions in specialized publications.
• Use precise technical vocabulary, since Princeton's GEO study shows that using precise technical terms can boost visibility by up to 28%. It helps AI match your content to specific queries.

Source Search Engine Journal | GEO study (arXiv) | Bing Blog

Here is a 5-step process to turn AI Performance data into concrete actions:

• Initial audit. Log into Bing Webmaster Tools and analyze your citation baseline over the past 30 days. Note total citations, number of cited pages, and the main grounded queries.
• Opportunity identification. Cross your grounded queries with your existing content. Spot queries where you are cited but content could be reinforced, and queries close to your expertise where you do not yet appear.
• Targeted optimization. For each identified page, apply the 7 strategies above. Prioritize already-cited pages (reinforcement) and create content for the detected gaps.
• Bi-monthly tracking. GEO optimizations propagate in 2 to 4 weeks per Princeton researchers. Analyze your metrics every two weeks and document the correlations between your actions and the results.
• Cross-platform reporting. Combine Bing AI Performance data with your ChatGPT monitoring tools (Profound, Hikoo, Otterly) and Google Search Console to get a 360° view of your AI visibility.

The AI Performance report is a public preview and comes with several limitations to keep in mind:

• No click data. The report measures citations, not the actual visits generated. Calculating an "AI click-through rate" is not yet possible.
• Sampled data. Microsoft notes that the data represents a sample, not the full citation activity.
• No API access. The data is not yet exportable through API, limiting reporting automation.
• Limited history. The report offers about 3 months of history (since November 2025).
• No segmentation by AI response type. It is impossible to distinguish citations coming from Copilot, Bing Chat, or integrated partners.

Microsoft is keeping its promises. Advanced segmentation landed in June 2026 with intents, topics, citation share, and the Compare view. Click metrics and API access are still expected during 2026. This is a tool to watch very closely.

The launch of AI Performance in Bing Webmaster Tools marks a turning point for the SEO industry. For the first time, content publishers have an official, structured tool to measure their visibility in AI-generated responses.

This report confirms that Generative Engine Optimization (GEO) is now a discipline in its own right, with its own metrics (citations, grounded queries, cited pages) and its own strategies. Brands that integrate AI Performance tracking into their SEO routine today will gain a significant head start.

At Hikoo, we integrate Bing AI Performance data into our multi-AI monitoring platform to give you a complete view of your GEO visibility on ChatGPT, Copilot, Perplexity, and Google AI Overviews. Because the future of search is multi-platform.