Bing's New AI Performance Report: How to See Exactly How AI Tools Find Your Business

Key Takeaways

• Bing Webmaster Tools has a free beta feature called the AI Performance Report that shows you how AI tools like Copilot discover and cite your content.
• Grounding queries are the actual search terms AI tools use behind the scenes to retrieve your website — they are different from regular user search queries and are now visible to you for the first time.
• You can access this data by setting up Bing Webmaster Tools (free, five minutes), then navigating to the AI Performance tab and scrolling to the Grounding Queries section.
• For each grounding query, search Bing manually — if you are not in positions one to three, that page needs work: tighten up the title, H1, URL, opening paragraph, and meta description.
• Improving your rankings for grounding queries follows the same playbook as regular SEO: internal linking, external backlinks, and more thorough content coverage.
• If Bing’s AI tools are citing you, there is a strong chance ChatGPT and other LLMs are too — they all use similar retrieval methods, so gains here compound across the board.

---

Most business owners have no idea how AI tools actually find and recommend their website. They assume it is some kind of magic, or that the AI just “knows” about them. The reality is far more mechanical — and far more useful for your business if you understand it.

Bing has just made that process visible. Their new AI Performance Report inside Bing Webmaster Tools shows you the exact search queries that AI tools run in the background when they pull content from your site to include in a response. This is genuinely new data that has never been available before, and right now it is free to access.

Here is what it is, how to read it, and what to do with it.

What the AI Performance Report Actually Is

The AI Performance Report is a beta feature inside Bing Webmaster Tools. It is designed to show you how Bing’s AI-powered products — things like Microsoft Copilot and Bing AI — interact with your website when they are generating responses for users.

Think of it like a regular performance report in Google Search Console, but instead of showing you how humans found your site through search, it shows you how AI tools retrieved your content to cite in their answers.

This is significant because it gives you a window into a process that was previously a black box. You can now see which pages are being used, what queries triggered those pages to be selected, and — critically — where you might be missing out on citations simply because your content is not ranking well enough for those specific queries.

The report is in beta, which means it is still being refined, but the core data it surfaces is already actionable.

What Are Grounding Queries?

This is the term Bing uses, and it is worth understanding properly because it changes how you think about AI search.

When someone asks Copilot or another AI tool a question, the AI does not rely purely on its training data to answer. Instead, it runs a set of real-time searches in the background — these are called grounding queries — to retrieve current, relevant information from the web. It then uses that retrieved content as the factual basis for its response, often citing the source.

So a user might ask: “What is the best web design agency in Auckland?” The AI might quietly run a grounding query like “top web design agencies Auckland NZ” or “Auckland web design company reviews” in the background. Whatever ranks well for those queries is what gets surfaced and cited.

The critical difference between grounding queries and regular search queries: grounding queries are generated by the AI itself, not typed by a human. They tend to be more literal, more topical, and more focused than how people naturally phrase questions. They are often closer to how you would write a research query in a library database than how you would ask a friend for a recommendation.

Bing Webmaster Tools is now showing you exactly which grounding queries led to your content being retrieved. That is the data you need.

How to Access the AI Performance Report

Step one: Set up Bing Webmaster Tools.

If you have not done this yet, go to webmaster.bing.com and sign in with a Microsoft account. You will need to verify ownership of your website, which you can do by adding a meta tag to your homepage or uploading an XML file. The whole process takes around five minutes, and if you already have Google Search Console set up, Bing lets you import your site from there automatically.

Step two: Navigate to the AI Performance tab.

Once you are inside Bing Webmaster Tools and have selected your site, look for the AI Performance section in the left-hand navigation. This is the new beta feature — it may not be in the same position for everyone while it is still being rolled out, so if you do not see it immediately, check for a “Reports” or “Insights” section.

Step three: Find the Grounding Queries section.

Within the AI Performance tab, scroll down until you find the Grounding Queries table. This is where the useful data lives. You will see a list of queries alongside metrics that show how often those queries led to your content being retrieved and cited in AI responses.

Export this list. You are going to work through it.

How to Use the Grounding Query Data

Now for the part that actually moves the needle.

Take your list of grounding queries and search for each one manually on Bing. Do this in a private or incognito browser window to avoid your personal search history influencing the results.

What you are looking for: are you appearing in positions one, two, or three for that query?

If your answer is yes, great — your content is well-positioned for retrieval. If the answer is no, you have found a specific, actionable gap. The AI is trying to use your content for that topic, but your page is not ranking well enough for it to reliably get picked.

For each page that needs improvement, go through this checklist:

• Title tag — does the grounding query language appear naturally in your title?
• H1 heading — same question.
• URL slug — is it clean and relevant to the topic?
• Opening paragraph — does the page answer the implied question within the first 100-150 words?
• Meta description — does it summarise the page topic clearly and include relevant terms?
• Topic coverage — does the page actually cover the topic comprehensively, or does it only touch on it?

That last point matters more than people realise. AI tools are not just looking for keyword matches — they are looking for pages that thoroughly address the topic they are researching. A thin page that mentions a subject briefly is going to lose out to a detailed page that covers it properly, every time.

How to Improve Your Rankings for Grounding Queries

Once you know which pages need to rank better, the approach is not fundamentally different from regular SEO — but the intent is more targeted.

Internal linking. This is often the quickest win. If you have stronger, more authoritative pages on your website, add two or three links from those pages to the page you are trying to lift. Use natural anchor text that includes the grounding query language — not forced, but relevant. Internal links pass authority and signal to search engines (and AI retrieval systems) what a page is about.

External backlinks. Links from other websites to that specific page will help it rank for competitive grounding queries. If you have pages that are borderline — appearing in positions four to ten — even a handful of quality backlinks can push them into the top three where AI retrieval becomes much more consistent.

Expand the content. If your page is 400 words and the competing pages are 1,200 words covering the topic in depth, you know what to do. Add sections that address related questions, include practical examples, and make the page genuinely the most useful result for that query. Comprehensive content does better in AI retrieval because it gives the AI more relevant information to work with.

The Cross-LLM Benefit

Here is something worth understanding about how the major AI tools work: ChatGPT, Perplexity, Gemini, and others all use broadly similar retrieval approaches. They retrieve content from pages that rank well in search for a given query.

That means if you are showing up consistently in Bing’s AI tools, you are very likely showing up in other AI tools too. The optimisation work you do based on Bing’s grounding query data is not just a Bing play — it compounds across every AI product that uses web retrieval to ground its responses.

This is why the AI Performance Report is worth your time even if you think Bing is not a major source of traffic for your business. The data it surfaces tells you something real about how AI tools in general are finding and using your content.

Pair This with Google Search Console

Bing Webmaster Tools gives you the AI-side picture. Google Search Console — particularly its new AI filter in the Search Type dropdown — gives you the Google side. Used together, you get a comprehensive view of how your content is being found and cited across both traditional search and AI-generated responses.

The workflow is simple: use grounding query data from Bing to identify pages that need work, make the improvements, and then monitor both Bing’s AI Performance Report and Google Search Console to track how those changes affect your visibility across both platforms.

We covered how to use the Google Search Console AI filter in detail in this post — if you have not read it yet, it is worth going through alongside what is covered here.

Start With What You Can See

Most of the work in AI search optimisation is still guesswork because the data simply does not exist. The AI Performance Report in Bing Webmaster Tools is a rare exception — it shows you something real about a process that has been completely opaque until now.

If you have not set up Bing Webmaster Tools yet, do it this week. It takes five minutes and the data you get back is more specific and actionable than most of the AI search advice you will find anywhere else.

If you want help working through the data and turning it into a concrete optimisation plan for your website, get in touch with the team at Lucid Media. We work with New Zealand businesses on exactly this kind of thing — making sure your site is visible not just in traditional search, but in the AI tools your customers are increasingly using to find services like yours.