Bing Webmaster Tools enriches its AI performance report with four new features

An already existing, but limited, AI performance report

Launched earlier this year, the Bing Webmaster Tools AI Performance Report answered a fundamental first question: Where is my content cited in AI-generated responses? It was a useful first step, but insufficient in the face of the increasing complexity of AI response systems.

These systems do not work like a traditional search engine. They synthesize information from several sources at the same time, in very varied contexts. A simple count of citations is therefore no longer enough to understand its real visibility.

It is to meet this need that Microsoft is today introducing four new capabilities, available in global preview. Features presented on the Microsoft Bing Blog by Krishna Madhavan, Meenaz Merchant, Saral Nigam and Trishna Shah.

Intents: understand why content is cited

The first new feature, Intents, classifies grounding queries (the queries that form the basis of AI responses) into broad categories: Informational, Commercial, Navigational, Learn and Solve, Research, Creation, Local, and others.

As a reminder, grounding designates the sources and web data that the AI ​​system uses to support and cite its answers. Until now, publishers saw which queries triggered a citation, without understanding what type of AI experience their content was engaged in.

With Intents, an e-commerce publisher may, for example, find that its content is mostly cited in comparison or purchase contexts, while an educational publisher may discover that its articles are mostly surfaced in “search” or “learning” type interactions. This information allows you torefine the structure and depth of content based on the types of AI experiences where they appear.

Topics: see your visibility by theme, not by keywords

The second feature, Topics, brings together grounding queries into larger thematic clusters. Where traditional SEO reasons query by query, AI systems reason by concepts and themes.

For example, queries like “ solar panels “, ” solar energy efficiency ” And ” residential solar installation » could all be grouped under a thematic cluster « Solar energy “. Publishers thus obtain an overview of the major themes that generate AI visibility for their site.

This approach is particularly useful for editorial teams, who naturally think in terms of themes and editorial territories rather than isolated keywords. Topics helps identify emerging areas of authority, spot gaps in topic coverage, and better understand how AI systems semantically group content.

Microsoft specifies that these classifications are based on evolving AI/ML systems. During the preview phase, some labels may still be quite broad, especially for very specialized areas. Accuracy is expected to increase over time.

Citation Share: measuring your share of presence in the AI ​​ecosystem

The third functionality provides comparative dimension. Citation Share calculates the percentage of citations attributed to a given site, compared to all the citations displayed for the same grounding query.

In short: in addition to knowing if you have been cited, you can know how much of the citation space you took up for this query. A site can be cited regularly but only represent a tiny fraction of the sources used on a subject. Citation Share makes this reality visible.

Microsoft insists that this indicator is intended as an observation metric, not as a ranking system or competitive score. It does not expose competing domains, does not represent a share of traffic, and does not assign a quality rating to content. Its goal is to allow publishers to identify areas where their representation in AI responses is strong or, conversely, fragmented across many sources.

Citation patterns can evolve for many reasons: changes in user behavior, evolution of AI models, freshness signals, partner update cycles, and broader changes to the web.

Compare: observe the evolution of citations over time

The fourth feature, Compare, allows you to overlay a previous period on the current reporting view. The goal is to easily visualize how citation activity is evolving, and relate these variations to content updates, seasonal effects, changes in demand, or broader ecosystem developments.

For example, publishers can compare the last 30 days with the previous 30 days, or define custom date ranges to analyze longer trends.

Compare does not seek to explain the causes of variations, but to make them visible in a clear and straightforward way, so that publishers can form their own hypotheses and adjust their content strategy accordingly.

What these tools actually change for publishers

Taken together, Intents, Topics, Citation Share and Compare mark a shift in approach in how Microsoft equips publishers with AI. Rather than reducing AI visibility to a single score or simple count, these features offer a set of complementary signals :

• The context of the quotes,
• Their thematic organization,
• Their relative weight
• Their temporal evolution.

For GEO practitioners, this data can be directly used to guide editorial decisions : on which topics to strengthen authority, in which types of AI queries to position yourself better, and how to monitor the impact of content adjustments over time.

These four features are available today in global preview in Bing Webmaster Tools. Microsoft also invites users to share their feedback directly from the dashboard interface via a new built-in feedback feature.