Can Bing's AI Performance report show how often competitors are cited?

Bing's AI Performance report, in public preview since February 2026 and expanded in June 2026, reports on the site you have verified in Bing Webmaster Tools. It does not list other sites, and it does not tell you how often a named competitor was cited.

Citation Share is the metric people expect to fill that gap. It doesn't. Microsoft calls it "an observational metric" and says it is "not a ranking system or a competitive scoreboard." Microsoft also states that it "does not expose competitor domains, represent traffic share, or assign quality scores to content."

So the report tells you how your site sits inside a pool of AI citations. It does not tell you who else is in that pool. If your goal is "are we cited more than Competitor X," this report cannot answer it, as of October 2026.

The next move is to use the report for what it is good at (your own citation trend and the queries behind it) and use a separate method for competitor comparison. Our guide to reading Bing's AI Performance report covers each metric in turn.

As of October 2026, the AI Performance report exposes these views for your own site:

• Total Citations: how often your pages were cited.
• Average Cited Pages: how many of your pages get cited on average.
• Page-level citation activity: which URLs are being cited.
• Grounding queries: the phrases the AI used when retrieving content it then cited.
• Intents and Topics: groupings such as informational, commercial and navigational.
• Citation Share and Compare: your share of citation activity, and comparisons over your own data.

Coverage spans Microsoft Copilot, AI-generated summaries in Bing, and select partner integrations. It is not a ChatGPT report, a point covered in whether the report includes ChatGPT citations.

The grounding queries are the most valuable part. Every prompt tracker tests questions its users chose. Grounding queries show the searches Copilot actually ran. You can use them later to pick better prompts for competitor checks, as described in tracking Copilot grounding queries.

A share is a fraction. Your citations sit on top, and all citation activity in the comparison set sits underneath. The bottom number is dominated by whoever already wins those queries, and you can't see it.

Here is a hypothetical. Say your site earned 200 citations last month and 300 this month. If the total pool grew from 2,000 to 4,000 over the same period, your share fell from 10% to 7.5%. Your absolute visibility grew by half. Your share still dropped.

That is why a falling share is not, on its own, a sign of lost ground. Read it next to Total Citations and page-level activity before you act. The difference between the two metrics is explained in citation rate vs citation share, and what a falling citation share means walks through the common cases.

The practical rule: treat Citation Share as context. Treat Total Citations and the pages behind them as the signal you report upward.

Since no first-party report names competitor citations, you need a proxy. The workable one is a prompt-based check.

1. Pick buyer questions. Start from the grounding queries in your Bing report, then add the questions your sales team hears.
2. Ask each engine the same questions. Record which brands each answer names and which sites it cites.
3. Score each engine separately. Engines draw on different indexes, so one blended number hides the differences.
4. Pick like-sized competitors. Brand prominence is the strongest association with being cited. A 2025 study of AI search citations (arXiv:2507.05301) found the top 20 news sources accounted for 67.3% of all citations for OpenAI models. If you benchmark a mid-market brand against a household name, you are mostly measuring brand size.
5. Repeat on a schedule. A single run is a snapshot. Trends need repeated runs with the same prompts.

Be plain about the limits. OpenAI says ChatGPT search typically rewrites a question into one or more targeted queries before it searches, so your prompt is not the query the engine runs. Answers also mix responses given with and without live retrieval. A prompt check is a proxy, not a measurement of the engine's internal ranking.

The free checker on this site runs up to three buyer questions and shows which brands get named, which is a reasonable first snapshot. For a longer walkthrough, see this free way to see which competitors AI recommends.

• Sampling. Microsoft labels the grounding-query data as a sample of overall citation activity. That caveat sits on grounding queries specifically. Details are in whether the data is complete or sampled.
• No API yet. As of October 2026 there is no public API for the AI Performance report. You export or record it by hand. See whether an API exists.
• Prompt counts undercount. In Otterly's 2,000 AI blogs experiment, Microsoft's AI Performance report credited one test site with 42,817 citations, while Otterly's prompt-based tracker logged 3,934. The time windows may not match, so don't read it as a ratio. It does show that a prompt sample sees less than the engine's own count.
• Engines diverge. In the same experiment, Copilot produced 69% of one site's tracked citations while ChatGPT cited it almost never. A competitor who leads in Copilot may trail elsewhere, and the reverse.