Microsoft today added a beta feature called Topics Portfolio to the AI Performance report in Bing Webmaster Tools, grouping the grounding queries that cite a website into broad themes and, for the first time in that report, listing queries where other sites are cited and the site in question is not.

In Short

Bing now sorts the questions people ask its AI into topics and shows a website which topics it gets quoted in. It also shows the questions in those topics where other websites get quoted instead, so you can see where you are missing. The numbers are rough estimates in a test version, and small sites may not see the feature at all.

What Microsoft shipped

The post, dated October 13, 2026 and published on the Bing blog under the Webmaster section, is signed by Krishna Madhavan, Saral Nigam and Vivek Jain of Microsoft Bing. It describes Topics Portfolio as a new tab inside AI Performance, sitting next to the existing Citations & Pages view. According to Microsoft, the feature "groups the grounding queries where your site appears into broader thematic clusters - for example, football, taxes or AI Chatbots - so you can see the shape of your presence across those clusters rather than one query at a time."

AI Performance itself is not new. Microsoft placed the dashboard into public preview on February 10, 2026, and PPC Land covered that launch when it gave publishers total citation counts, average cited pages per day, a sample of grounding queries and page-level citation activity across Microsoft Copilot, AI-generated summaries in Bing and select partner integrations. Six weeks later, on March 23, the dashboard gained a two-way mapping between grounding queries and pages, so that one query could be traced to several URLs and one URL to several queries.

Everything in those earlier versions was confined to a site's own activity. A publisher could see which of its pages were cited and for which queries, but not how large a theme was across the web, how many other domains competed for it, or which queries went to someone else. Topics Portfolio is the first release in the report to bring that ecosystem data in.

Microsoft frames the addition as the "broader view" on top of query-level data: "how that presence is distributed across themes, how much demand each theme represents, how competitive each theme is and where your site has room to grow."

The landing page and its four columns

Access depends on volume. According to Microsoft, a site will only see the new tab "if there is sufficient data." When it does appear, the landing page opens to a personalised list of topics where the site has what the post calls "a meaningful presence," with three ecosystem measures beside each one.

The four columns are defined in the post as follows:

  • Topic - a broad thematic cluster of related grounding queries where the site appears.
  • Topic Size - the approximate total citations across all sites for the topic over the selected period, which Microsoft describes as "a measure of how much AI Search demand the topic represents."
  • Sites Cited - the approximate number of sites appearing for queries in the topic, an indicator of how crowded it is.
  • Queries - the approximate number of distinct grounding queries in the topic.

Each of the three numeric columns carries the word "approximate" in Microsoft's own definition. That matters, because Topic Size is the closest thing the report has offered to a demand estimate. It counts citations, not searches or users. A topic with a large Topic Size is one where AI answers cite sources often, which may reflect how many queries arrive, how many sources each answer pulls in, or both. The post does not separate those effects.

Microsoft is explicit that the list is not generic. "The topics and the opportunity queries are surfaced from your own presence and activity in AI Search, not from a generic ecosystem list," according to the post. Two sites in the same vertical could therefore see different topic labels and different boundaries for what looks like the same theme.

Top Opportunity Queries

Below the topic table sits a second list, Top Opportunity Queries. Microsoft defines these as "the strongest queries across your whole portfolio where other sites appear and yours does not." It calls the list "a fast, cross-topic read on where you could grow."

Citation volumes for these queries are not shown as exact figures. According to Microsoft, they appear "as ranges rather than exact numbers." The post later describes this as deliberate. It does not say how wide the ranges are or how they are bucketed.

The post does not define "strongest." It could mean the highest citation volume, the largest gap between the site and others, or some other ranking. Microsoft also does not say how many opportunity queries are shown per topic or across the portfolio.

Inside a topic

Clicking a topic opens a detail page. Three KPI cards sit at the top: Topic Size, the site's own citations within the topic, and the number of queries in it. Beneath the cards, the queries are split into two tables.

Your Grounding Queries lists the top queries in the topic where the site already appears. Microsoft says these are "aligned with the data you already know from Citations & Pages," which suggests the counts should reconcile with the query-level data that has existed since February, though the post does not state that they will match exactly.

Opportunity Queries lists the top queries in the same topic where other sites appear and the site does not, again with citation volumes as ranges. "Together they give you the two halves of a topic: what you already own, and what is still open," according to Microsoft.

The detail page creates a ratio that was not previously visible in Bing's tooling. A site can now set its own citations against the topic's total. The post does not compute this as a share figure, and it does not label anything as share of citations or share of voice. With the topic total marked as approximate and opportunity volumes reported in ranges, any share calculated by hand would carry the same imprecision.

Who Microsoft says it is for

According to Microsoft, Topics Portfolio "is aimed at anyone who wants to understand and grow their site's presence in AI-generated answers," and the company expects it to be "most useful to publishers doing content strategy, editorial planning and competitive positioning."

The post lists three things the report "could not show before":

  • A portfolio view - "the thematic shape of your AI Search footprint at a glance."
  • Ecosystem context - "how big each topic is, and how crowded it is with other sites."
  • Actionable opportunity - "specific queries within specific topics where you can grow, drawn from real grounding activity happening on Bing."

The phrase "real grounding activity" ties the feature back to grounding, the process by which an AI system retrieves documents to support a generated answer. Bing's grounding index serves Copilot and AI summaries, and in May 2026 Microsoft published a technical post arguing that the index serving AI answers is evaluated on different criteria from the one serving ranked links. PPC Land reported on that post, which named factual fidelity, source attribution, freshness, coverage of high-value facts and contradiction handling as the dimensions where the two diverge.

The beta caveats

Microsoft devotes a full section to limitations, and the language is unusually direct for a product post. "Treat the numbers as directional signals for planning, not as a precise ledger," it states. The company says it would "rather put it in front of you early and improve it with your input than hold it back."

Four limitations are listed.

Classification can be wrong. According to Microsoft, topic classification "is AI-generated and can occasionally be imperfect, so a query may land in a topic you would have labelled differently." The post does not name the model that performs the classification. Microsoft Clarity's Topic Insights feature, launched in July, runs on GPT-5.3, as PPC Land reported at the time; there is no statement linking the two products' classification systems.

Small sites may be excluded entirely. "Topics with insufficient site presence or insufficient ecosystem activity are not shown, so smaller sites may see a shorter list or may not even see this feature," according to Microsoft. No threshold is published. A site cannot know in advance whether it qualifies, and a missing topic does not mean the site has no citations there.

Branded queries can appear as opportunities. Microsoft's own example uses its fictional Contoso brand: "In a Running Shoes topic, for instance, contososhoes.com may see Air Shoes listed as an opportunity even though it is a Compete-branded query." In other words, the opportunity list may contain queries about a competitor's products that the site has little realistic chance of being cited for. The post offers no filter to exclude competitor-branded queries.

All ecosystem figures are approximate. Topic Size, Sites Cited and Queries are approximate, and opportunity citation volumes are deliberately shown as ranges, according to the post.

None of the caveats address sampling directly. When AI Performance launched, Microsoft described grounding queries as a sample of citation activity, and PPC Land's March coverage noted the data was not a complete log. The new post does not say whether ecosystem totals such as Topic Size are drawn from that same sample or from a fuller count.

A small naming inconsistency

The post refers to the new tab as "Topics Portfolio" in its main description, saying it appears "next to Citations & Pages." In the closing "Getting started" section, it instructs users to "select the Topics tab." Whether the tab is labelled differently in some accounts, or the closing instruction is shorthand, is not stated.

Feedback built into the rows

Every topic row and every opportunity query carries a thumbs up and thumbs down control, according to Microsoft. "If a topic does not describe your content, or an opportunity query is off the mark for your site, mark it down," the post states. "Those signals go straight to the team working on the classification and the opportunity selection, and they shape what we fix first."

The mechanism means that publishers using the beta are also labelling training and evaluation data for Microsoft's classifier. The post does not say whether a downvote changes what that particular site sees, or only informs future model changes. Users can also report issues through the usual Bing Webmaster Tools feedback channels.

Date ranges

The feature offers a one-, two- or three-month window, according to the "Getting started" section. That is shorter than the historical retention Bing Webmaster Tools offers for traditional search data, which Microsoft extended to 24 months in August 2025. A three-month ceiling limits how far back a publisher can compare topic sizes, and it means seasonal topics may appear or disappear from the portfolio depending on when the report is opened.

How this compares with other AI visibility tools

Microsoft now runs two free products that report on AI citations, and they have been converging. Microsoft Clarity's Citations dashboard reached general availability on May 13, 2026, reporting page citations, share of authority, AI referral traffic and grounding queries. Clarity then added Topic Insights on July 9, a Query Topics beta that groups grounding queries into generated topics on July 22, branded and non-branded segmentation on August 3, and on September 29 the ability for brands to edit which terms count as branded.

Topics Portfolio overlaps with that work. Both products now cluster grounding queries into topics. The difference lies in what sits beside the clusters. Clarity reports share of authority, a ratio of a domain's citations to all citations. Bing's new view reports raw topic totals, the number of competing sites and a list of queries the site is missing, but no share figure. Clarity also gained a branded query filter in August, which is precisely the gap Microsoft acknowledges in Topics Portfolio's opportunity list. The two teams have not explained publicly why the same company maintains two overlapping citation dashboards with different metrics.

The Bing blog lists a related post dated June 16, 2026 titled "New AI Visibility Insights in Bing Webmaster Tools: Intents, Topics, Citation Share, Compare." Its title names topics and citation share, which suggests some form of topic grouping and share reporting already existed in Bing Webmaster Tools before today. Today's post does not reference that earlier release or explain how Topics Portfolio differs from it.

Google's reporting is thinner. Search Console's generative AI performance report launched on June 3, 2026 for a subset of sites, covering AI Overviews and AI Mode with impressions, pages, countries, devices and dates. By the end of August it had reached every site, but without click or query data. Google offers no equivalent to grounding queries, let alone a list of queries where competitors are cited instead.

Why it matters for marketers

For most of the past year, measuring visibility in AI answers has depended on third-party tools that run synthetic prompts against chatbots and record which brands appear. Semrush's AI Visibility Index, for example, was built on 126 million US prompts collected from January to April 2026, and its accompanying survey found 45% of 481 marketers said they could not properly measure AI answer visibility. Those tools estimate exposure from the outside.

Topics Portfolio is the search engine itself describing citation activity on its own AI surfaces, drawn from what Microsoft calls real grounding activity. That is a different kind of evidence from a panel of prompts. It is also confined to Bing's surfaces, which hold a small share of search. Bing's US share peaked at 5.33% in December 2025 before falling to 3.2-3.8% in early 2026, according to Datos figures reported by PPC Land in June.

The opportunity list is the more consequential part. Until now, Bing's AI Performance data could tell a publisher what it was winning. It now also names specific queries that are going to other domains, which is the raw input for the kind of content planning that GEO practitioners have been trying to reverse-engineer from outside tools. Whether those queries reflect durable demand is harder to judge. Volumes are ranges, topic sizes are approximate, competitor-branded queries are mixed in, and smaller sites may be shown nothing at all.

There is also a question about how much the feature discloses about competitors. Sites Cited is a count, not a list of names, and opportunity queries say that other sites are cited without saying which ones. Microsoft has kept the competitive data aggregate. A publisher learns that a topic is crowded and that a query is going elsewhere, but not to whom.

Timeline

Summary

Who: Microsoft Bing, through a post by Krishna Madhavan, Saral Nigam and Vivek Jain, for website owners and publishers using Bing Webmaster Tools.

What: A beta feature called Topics Portfolio inside the AI Performance report. It groups a site's grounding queries into AI-classified topics, shows approximate Topic Size, Sites Cited and Queries for each, and lists opportunity queries where other sites are cited and the site is not, with volumes given as ranges.

When: Today, October 13, 2026, building on the AI Performance public preview from February 10, 2026.

Where: In Bing Webmaster Tools, covering citation activity on Microsoft's AI surfaces such as Copilot and AI summaries in Bing, with one-, two- or three-month date windows.

Why: Microsoft says the view adds ecosystem context and specific growth opportunities to query-level citation data. The figures are approximate, classification can be wrong, competitor-branded queries can appear as opportunities, and sites without enough data may not see the feature.