Microsoft has added AI Search Queries to a new Topics Portfolio tab in the beta AI Performance report for Bing Webmaster Tools, according to the Topics Portfolio beta announcement. The view groups queries that lead to citations in AI answers, then highlights queries where other sites appear and the site being reviewed does not.
Bing calls those terms grounding queries. The company says its AI uses them to find and cite content in Copilot, Bing’s AI summary and select partner integrations. The feature appears only when Bing has enough data about a website, so smaller sites may see fewer topics or none.
What AI Search Queries Reveal
The Topics Portfolio tab sits to the right of Citations & Pages in AI Performance. Its first screen lists topics where Bing considers a site to have a “meaningful” presence. Each topic has three estimated measures: Topic Size, Sites Cited and Queries.
Topic Size estimates the total citations across all sites for a topic over the selected period. Sites Cited approximates how many sites appear for the topic’s queries, while Queries counts the different grounding queries grouped there. The figures describe estimated activity, not an exact count of a site’s results.
Below the topic table, Top Opportunity Queries gathers terms across topics where other websites receive citations and the reviewed site does not. Bing displays opportunity volumes as ranges. Selecting a topic opens its size, the site’s citations and query count, plus separate tables for queries where the site appears and opportunities where it does not.
Users can view the data over one-, two- or three-month windows. The report also offers thumbs-up and thumbs-down feedback on each opportunity query or topic row. Microsoft says those ratings go to the team responsible for classification and opportunity selection, helping shape what it addresses first.
How the Beta Differs From Earlier AI Search Queries
The June update grouped grounding queries where a site already received citations. Microsoft says Topics Portfolio now adds a total across all sites at the topic level and surfaces queries where the site is absent. AI Performance entered public preview in February; the June update also included Topics, Intents, Citation Share and Compare.
Citation Share, introduced in June, shows a site’s percentage of citations for a grounding query. Bing says the measure does not reveal competitor domain names or their data. It is intended to show a site’s presence relative to others.
Estimates Need Editorial Review
Microsoft says artificial intelligence generates the topic labels, so a query may land under a label that differs from how a site owner would classify it. Branded searches can also appear as opportunities. Microsoft gave the example of a shoe website seeing a competitor’s brand search listed under Running Shoes.
“Treat the numbers as directional signals for planning, not as a precise ledger,” Microsoft wrote. Because topic measures are estimates and opportunity volumes appear in ranges, the report supports approximate comparisons rather than tracking exact citation counts. Teams can review each query’s fit before treating it as a content gap.
At the time of the report, Bing’s AI Performance help page did not cover Topics Portfolio, leaving Microsoft’s limits section as its main written guidance. Microsoft had not given a date for the beta to end. For now, teams can inspect the available query examples, use the feedback controls when classifications miss the mark, and treat the estimates as planning context. The broader question of how SEO benefits reach beyond rankings is also relevant as teams assess visibility across AI Search Queries.
For related context, readers can review Google’s position on audio and its separate guidance that domain endings bring no bonus. Microsoft’s beta gives site owners another way to inspect where AI answers cite their pages, while its estimates and classification limits call for human review.
