Microsoft launched an AI Performance dashboard in Bing Webmaster Tools this week. For the first time, brands can see hard numbers on how often their content gets cited in AI answers from Copilot, Bing AI summaries, and partner integrations.

This matters because every other major AI platform treats visibility as a black box. ChatGPT doesn’t tell you when it referenced your product. Google doesn’t show you citation data from AI Overviews. Perplexity doesn’t share metrics on source attribution. Microsoft is the first to give brands actual analytics on AI visibility, not just traditional clicks and impressions.

What Microsoft Actually Released

The AI Performance report shows five metrics that matter:

Total citations: How many times Copilot and Bing AI referenced your content as a source. This is your “AI visibility” score.

Average cited pages: How many unique pages from your site get cited per day on average. This shows breadth, not just volume.

Which pages get referenced: The specific URLs AI systems pull from when they cite your brand.

Grounding queries: This is the important one. These are the phrases the AI actually used behind the scenes to retrieve your content, not the words users typed.

When a user asks “best running shoes for marathon training,” the AI might use grounding queries like “marathon shoe cushioning stability” or “long-distance running footwear performance” to find sources. If those phrases don’t exist on your product pages, you have an optimization gap.

Page-level trends: How citation volume for individual pages changes over time.

Microsoft notes the grounding query data is a sample, not a complete log. You’re seeing patterns, not everything the AI did.

Why Grounding Queries Matter

You can’t optimize for what users type because AI doesn’t search for what users type. It rewrites their question into multiple retrieval queries behind the scenes.

If users search for “natural skincare for sensitive skin” but the AI uses grounding queries like “hypoallergenic facial products ingredient list” and “dermatologist-tested sensitive skin formulations,” your content needs to cover both layers. The user-facing language and the technical, attribute-heavy language AI uses to actually find your pages.

This is fundamentally different from traditional keyword optimization. With SEO, you target the phrases users search for. With AI optimization, you’re anticipating how systems will decompose user intent into structured retrieval logic that may look nothing like the original question.

The grounding query data gives you visibility into that retrieval layer for the first time. If the phrases showing up in your dashboard don’t match what’s on your product pages, you know exactly what language to add.

Microsoft vs. Google on Transparency

Microsoft is giving brands citation data. Google is not.

Here’s why that matters. Google’s AI Overviews appear in a significant portion of search results and reach over 90% of global search volume. That’s where the business impact lives for most brands. But Google provides zero official analytics on which pages get cited, how often, or what retrieval logic it’s using.

Microsoft’s dashboard gives you visibility into citations, grounding queries, and page-level trends. That’s useful feedback. But Bing represents roughly 3-4% of search volume. The platform giving you transparency is not the platform driving most of your traffic.

This creates the central strategic challenge. Google is the big opportunity. That’s where brands need to focus AI optimization efforts because that’s where the volume is. But Google won’t tell you if your optimization is working. You can invest in structured data, clean product feeds, and content changes, but Google keeps AI citation data completely opaque.

For brands, this means using Bing’s AI Performance data as a proxy for what you can’t see on Google. If your citation volume on Bing is growing and the grounding queries make sense, there’s a reasonable chance similar dynamics are playing out on Google’s side. You’re testing on the platform that gives you feedback, then applying those learnings to the platform that actually matters for business outcomes.

What You Can Do Now

Even with incomplete data, the dashboard gives you levers to act on.

Check which pages get cited most often. If your product pages are getting cited but category pages aren’t, or vice versa, that tells you where AI systems are finding value in your content structure.

Review grounding queries for gaps. If the phrases AI uses to retrieve your content don’t match the language on your pages, you have a clear optimization target. Add the technical, attribute-heavy phrasing alongside the user-facing descriptions.

Track citation trends over time. If changes to product descriptions, structured data, or feed quality correlate with increases in citations, you’re getting signal on what works.

Layer in your own attribution. Connect AI citation volume to downstream metrics like branded search, direct traffic, and conversions from Bing. That tells you whether visibility is driving outcomes or just vanity metrics.

Use Bing as a testing ground. If you’re unsure whether a content change will improve AI visibility, test it on Bing where you can measure citation impact. If it works, apply the same logic to Google optimization even though you won’t get the same feedback.

The Practical Takeaway

Microsoft’s AI Performance dashboard is the first real analytics tool for AI visibility. It’s not complete, it’s not perfect, and it doesn’t solve attribution. But it’s the only platform giving brands hard numbers on how often their content gets cited in AI-generated answers.

The strategic opportunity is Google. That’s where over 90% of search volume lives, where AI Overviews reach the most users, and where brands need to focus optimization efforts for business impact. The problem is Google won’t give you the data to know if your optimization is working.

That makes Bing valuable not because of its own traffic volume, but because it’s your testing ground for strategies you’ll apply to Google. You test content changes on Bing, measure citation impact, and use those learnings to inform your Google optimization even though you can’t confirm the results there.

For brands already optimizing for AI discovery, Bing’s dashboard is the first feedback loop that shows whether your work is landing. For brands that haven’t started, this is a clear signal that AI visibility is measurable on at least one platform, and the optimization principles likely transfer to the platform that matters most.

If you haven’t assessed your AI visibility across platforms like Google, ChatGPT, and Bing, the tool we launched a few weeks ago covers that: https://commerce-gpt.yotpo.com

Track your visibility for free here: Commerce GPT Visibility Tool

The big opportunity is Google. Bing just happens to be where you can measure progress.

Tomer