Bing Webmaster Tools now measures AI-search citation share, intent and topic visibility

Microsoft expanded Bing Webmaster Tools’ AI Performance public preview on June 16, 2026 with four additional reporting dimensions: Intents, Topics, Citation Share and Compare. The original February preview showed total citations, cited pages, grounding queries and URL-level citation activity across Microsoft Copilot, AI-generated Bing summaries and selected partner integrations. The June expansion adds classification of the queries that lead to citations, thematic grouping, a relative citation-share measure and period-over-period comparison. On October 13 Microsoft launched Topics Portfolio (Beta) inside AI Performance. For eligible sites it groups grounding queries into personalized themes, adds topic-level approximate total citations across sites (Topic Size), number of sites cited and distinct queries, and surfaces Opportunity Queries where other sites are cited but the site is not. It adds a topic-level competitive and content-gap view beyond June's topic classification, citation-share and period comparison.

AI-search optimization has been difficult to evaluate because publishers often see referral traffic only after a click and receive little information about when their content was used inside an answer. Bing’s expanded reporting gives site owners a first-party way to measure citation visibility before the click: what kinds of user intent their pages satisfy, which topics they are associated with and whether citation share is improving or weakening. That does not prove downstream conversion, but it gives publishers a much better feedback loop for deciding what content to update, consolidate or expand. The new portfolio view gives publishers a concrete, first-party shortlist of topical gaps backed by Bing's observed grounding activity rather than generic keyword estimates. Small publishers should not confuse absence from the report with zero AI visibility: Microsoft suppresses topics with insufficient activity and shows opportunity volumes as ranges. No click or conversion attribution is provided.

The February AI Performance preview established a baseline of total citations, average cited pages, grounding-query phrases, page-level citation activity and trends. The June update adds intent and topic classification so publishers can see the context in which their content is being used, rather than treating every citation as equivalent.

Citation Share is designed to show a site’s relative presence in AI answers over time. It is not a conventional organic ranking metric and Microsoft does not present it as a direct proxy for authority, but it lets teams monitor whether their share of citation opportunities is changing within the supported AI experiences.

The new comparison view lets publishers contrast periods rather than manually exporting snapshots. For teams running content refreshes, schema changes or topic expansion, this makes it possible to evaluate whether citation patterns moved after the intervention, while still avoiding a false claim of causality.

The reporting does not represent the entire generative-search market. It reflects Microsoft Copilot, AI-generated Bing summaries and selected partner integrations, and Microsoft explicitly describes the feature as a public preview. Builders should therefore use it as one source of evidence alongside referral analytics, other AI-search tools and conversion data.

The new Topics Portfolio tab appears beside Citations & Pages only when a verified site has sufficient AI Performance data. Its portfolio table lists thematic clusters, approximate Topic Size (total citations across all sites), Sites Cited and distinct Queries. A Top Opportunity Queries view shows queries where other sites appear in AI-generated answers and yours does not; selecting a topic separates Your Grounding Queries from Opportunity Queries. Publishers can use these signals to prioritise coverage gaps, but the figures are directional, not precise market-share or traffic counts.

Microsoft says topic classification is AI-generated and may mislabel queries, smaller sites may see fewer or no topics, branded competitor queries may be labelled opportunities, and aggregate counts are approximate while opportunity citation volumes appear as ranges. Review one-, two- or three-month periods and corroborate topics against actual content expertise and analytics before changing editorial plans. This is Bing AI-grounding data, not proof of Google/ChatGPT coverage or downstream visits.