Microsoft Clarity Splits AI Citations

Microsoft Clarity has introduced new ways for marketers to understand how their websites appear in AI-generated answers, including separate reporting for branded and non-branded citations.

The update gives businesses a clearer view of whether AI systems are finding their content because users are researching the company specifically or because the content is relevant to a wider topic.

This distinction could become increasingly useful as businesses look beyond traditional search rankings and try to measure their visibility across AI-powered search and answer platforms.

What Has Microsoft Changed?

Clarity’s AI Citations dashboard now allows marketers to distinguish between branded and non-branded grounding queries.

A branded query is one where the AI system’s underlying search specifically refers to a company or brand. A non-branded query, meanwhile, relates to a broader subject, product category or problem without directly naming the business.

The dashboard now provides:

• Labels identifying branded grounding queries.
• Filters for comparing branded and non-branded activity.
• Citation totals for each category.
• Separate Share of Authority figures.

Previously, marketers could examine individual grounding queries, but comparing larger groups was more difficult. The new filters make it easier to identify where a website’s AI visibility is coming from.

What Is Share of Authority?

One of the key metrics included in the update is Share of Authority.

Microsoft uses this measurement to show the percentage of citations attributed to a particular website compared with other domains that were cited for the same included queries.

However, it is important not to interpret this as a website’s overall share of the AI search market.

Instead, it reflects the proportion of citations a domain received within the specific queries in which it appeared.

This makes the metric particularly useful for comparing performance against competing websites within particular subjects or categories.

For example, a company might have a strong Share of Authority for branded searches but a much weaker position when AI systems are researching a broader industry topic.

Branded Citations Show Brand Visibility

Branded grounding queries can provide insight into how often AI systems use a company’s own website when researching that business.

These searches could relate to areas such as:

• Product information.
• Pricing.
• Customer support.
• Product specifications.
• Company policies.
• Reviews and comparisons.

Strong branded visibility may indicate that an AI system considers the company’s own website a useful source when looking for information about the brand.

However, it does not necessarily mean that users are searching for the brand directly. The AI system may have introduced the company into its research process after receiving a much broader question.

Non-Branded Citations Show Wider Authority

Non-branded queries are potentially even more valuable for businesses trying to build authority within their industry.

These searches do not specifically mention the company. Instead, they focus on broader topics, problems or categories.

For example, an AI system could search for information about the best analytics platforms, ways to reduce customer acquisition costs or tools for understanding website visitors.

If a company’s content is cited in these searches, it suggests the website has become a source of information for the wider subject rather than simply for queries about its own brand.

This gives marketers an opportunity to see how their expertise compares with competing sources.

In simple terms, branded citations show where your company appears in AI research, while non-branded citations show where your expertise is being recognised across a wider topic.

Grounding Queries Are Not the Same as User Searches

One important point marketers need to understand is that a grounding query is not necessarily the same thing as the question originally asked by the user.

An individual might ask an AI assistant a broad question about choosing website analytics software. The AI system could then perform several additional searches, including searches for specific companies or products.

Clarity records those underlying searches rather than the user’s original wording.

As a result, a branded grounding query should not automatically be interpreted as evidence that the user already knew about or preferred that particular company.

Similarly, non-branded activity does not necessarily indicate that the user was at the beginning of the buying journey.

The data is best understood as showing AI retrieval behaviour, rather than providing a direct measurement of customer intent.

Businesses should combine Clarity’s information with other data, including website visits, user behaviour, enquiries, sales and CRM information, to understand the commercial impact.

Citations Do Not Automatically Mean More Traffic

Another important distinction is between AI citations and AI referral traffic.

A citation means that an AI-generated response referenced a particular website or page. It does not necessarily mean the user clicked through to that website.

In some cases, an AI answer may provide enough information for someone without them needing to visit the original source.

This means marketers should avoid assuming that an increase in citations will automatically result in more visitors, leads or sales.

A more useful measurement framework separates performance into three stages:

1. AI citation visibility – how frequently AI systems reference your content.
2. AI referral traffic – how many visitors arrive at your website from AI platforms.
3. Business outcomes – leads, sales, revenue or other conversions generated from those visitors.

Each measurement answers a different question, so combining them into one figure could give a misleading picture of performance.

A More Useful AI Visibility Scorecard

The new reporting features allow businesses to build a more detailed picture of their AI visibility.

Metric	What it shows
Branded citations	How frequently your content is cited for brand-related queries
Branded Share of Authority	Your share of citations within included branded queries
Non-branded citations	How often your content appears for broader subjects
Non-branded Share of Authority	Your share of citations within wider category or topic searches
AI referral traffic	Website sessions originating from AI assistants
AI-referred conversions	Leads, sales or other outcomes attributed to AI traffic

Reviewing these figures alongside the individual pages and topics receiving citations can make the data more useful.

For example, weak branded visibility could indicate that product pages, pricing information or company documentation need improvement.

Meanwhile, weak non-branded visibility could suggest that a business needs stronger educational content, original research or greater coverage across its industry.

How Marketers Can Use the Data

Businesses can use the new filters to establish a baseline and monitor changes over time.

A simple monthly review could involve comparing branded and non-branded citations, identifying which topics have changed, checking Share of Authority and then deciding whether content or marketing activity needs to be adjusted.

A fall in non-branded visibility, for example, could prompt a review of competitors that are receiving more citations.

On the other hand, growing visibility around a particular topic could indicate an area where additional content, research or digital PR investment could be worthwhile.

The key is to look beyond the headline citation number and understand which topics, pages and competitors are behind the change.

Clarity Data Has Limitations

While Microsoft’s new reporting can provide useful insight, marketers should remember that it does not represent every AI interaction across the internet.

Microsoft describes the AI Citations dashboard as a representative view of activity across supported AI experiences.

The data is designed to help businesses identify trends and make comparisons rather than provide a complete record of every AI-generated answer or citation.

Some low-volume activity may not appear, and figures can change as Microsoft processes additional information. Totals can also differ depending on whether the data is viewed by page or by grounding query.

The exact AI platforms covered by Clarity should also be considered when interpreting the figures.

For reporting purposes, businesses should therefore make it clear that their figures represent Microsoft Clarity AI citation data, rather than claiming they represent total AI visibility across every platform.

What This Means for Businesses

The introduction of branded and non-branded reporting gives marketers a more detailed way to measure how their content is being used by AI systems.

Instead of looking at one overall citation figure, businesses can now ask more useful questions:

• Is AI citing us because of our brand?
• Is our content being discovered for wider industry topics?
• Which competitors are receiving more citations?
• Which pages are being used most frequently?
• Is increased AI visibility resulting in website traffic?
• Are those visitors generating leads or sales?

These distinctions are important because visibility alone does not necessarily translate into commercial results.

For businesses investing in AI search visibility, the most effective approach is likely to combine citation data with referral traffic, website analytics and conversion information.

Key Takeaways

Microsoft Clarity’s latest update makes AI visibility reporting more useful by separating branded and non-branded citations.

Branded citations can help businesses understand how AI systems use their own websites when researching a company or its products. Non-branded citations provide a broader view of whether a business is being recognised as an authority within its industry.

However, grounding queries should not be treated as a direct representation of the user’s original search or buying intentions.

Most importantly, marketers should keep citations, AI referral traffic and conversions separate. A website can receive a large number of AI citations without seeing a similar increase in traffic or revenue.

As AI-powered search continues to develop, these distinctions will become increasingly important for businesses trying to understand where their online visibility comes from and whether that visibility is actually contributing to growth.