Microsoft Clarity Branded Queries Separate Brand Visibility From Discovery

Ankita Pathak Avatar
✨ Summarise and Analyse the Article

Microsoft Clarity has added branded vs non-branded segmentation to its AI Citations dashboard, giving marketers a clearer way to distinguish brand-specific retrieval from broader category discovery. The Microsoft Clarity branded queries update went live on August 3, 2026, adding branded labels to grounding queries, filters for branded and non-branded AI queries, and a Share of Authority breakdown by query type.

For B2B SaaS marketing teams, the useful part is not another dashboard filter. It is the ability to ask a more important question about AI search visibility: are AI systems retrieving your content only when your brand is part of the retrieval process, or can they also find you while researching the broader category, problem or use case?

If citation growth is concentrated in branded AI queries, the increase is being driven mainly by brand-specific retrieval. If non-branded AI queries grow, that can be a useful directional signal that your content is entering broader category research. The important caveat is that a branded grounding query does not prove the user originally mentioned or even knew the brand. Microsoft says grounding queries may differ from the user’s actual prompt. Microsoft’s Clarity documentation explains that distinction.

What changed in Microsoft Clarity

Microsoft’s August 3 announcement introduced three practical additions to the AI Citations dashboard:

  1. Branded labels on individual grounding queries, so marketers can see when the retrieval query references their company or brand.
  2. Filters for branded and non-branded queries, making it easier to compare brand-specific retrieval with broader category discovery.
  3. A branded vs non-branded Share of Authority split, so citation presence can be analysed separately instead of being compressed into one total.

The key term is grounding queries. These are retrieval queries AI systems use to find supporting information before generating an answer. They are not necessarily the words a user typed. For example, someone could ask an AI assistant for tools that help analyse website behaviour without mentioning a company, while the AI system generates brand-specific grounding queries as it assembles the response.

Query typeWhat it means in ClarityWhat marketers can infer
BrandedThe grounding query references your company or brandYour content is being retrieved in a brand-specific context, but this does not prove the user named the brand
Non-brandedThe grounding query focuses on a broader category, problem, feature or topicYour content is participating in broader discovery around the market or use case

What Share of Authority actually tells you

Clarity also splits Share of Authority by branded and non-branded query type. Microsoft defines the metric as the percentage of citations attributed to your domain compared with citations from other domains within the relevant calculation. The methodology matters: the calculation is applied daily, and when your domain receives a citation for a query on a given day, citations from all domains for that query-day are included. Query-days where your domain received no citation are excluded.

That means Share of Authority should not be read as your percentage of the entire AI search market. It is better used as a competitive signal showing how much citation presence your domain earns within the queries where it participates. The branded and non-branded split makes this more actionable because a company can now see whether its citation strength comes mainly from brand-specific retrieval or from broader category discovery.

4 ways the branded vs non-branded split helps marketers

1. It separates brand-specific retrieval from broader discovery

Paid search teams already know why brand and generic demand should not be blended into one number. Someone searching directly for your company behaves differently from someone searching for a category such as “customer onboarding software.” That is one reason a clean Google Ads campaign structure separates brand campaigns from in-market and other forms of intent.

The same reporting discipline is useful for AI citations. Strong branded retrieval shows that your content is being found in a brand-specific context, while strong non-branded retrieval suggests your site can also surface when the system investigates broader problems, categories or use cases. Non-branded citations are not a pipeline metric, but they can be a useful diagnostic for category discovery.

2. It gives SEO and content teams a better content-gap signal

A site can perform well for brand-specific retrieval while contributing very little to broader discovery. If branded queries dominate, content teams should inspect whether category pages, comparison pages, use-case content, product education and expert resources answer the broader questions that appear before a buyer settles on a vendor.

This is closely related to the challenge addressed in our guides to AI Search SEO and generative engine optimisation. The goal is not to repeat a keyword more often. It is to give search and AI systems enough useful, specific and credible information to understand where the company belongs within a category.

3. It helps separate visibility from traffic

Clarity’s Citations dashboard measures activity that can happen before a website visit. Microsoft says the dashboard includes page citations, grounding queries, cited pages, Share of Authority and AI referral traffic. Citation counts show how often a page was referenced, but they do not represent the page’s ranking or prominence within an AI-generated answer.

That distinction matters because more of the research journey can happen inside an AI-generated experience. Our analysis of Google’s shift toward agentic search explores the same change from the Google side. Once a visit does happen, the GA4 AI Assistant channel group can help separate recognised AI-referred sessions, while CRM and pipeline data determine whether that visibility created qualified business outcomes.

4. It gives different AI visibility signals different jobs

The practical mistake would be replacing one vanity metric with two vanity metrics. A better approach is to give each signal a specific role in the scorecard.

SignalWhat it showsUseful interpretationBest next action
Branded AI queriesGrounding queries that reference your brandBrand-specific AI retrievalCheck product, pricing, integration and proof pages for accuracy and consistency
Non-branded AI queriesBroader category, problem or topic retrievalDirectional category discoveryMap commercially relevant queries to cited pages and identify content gaps
Share of AuthorityYour citation share within included query-days where your domain appearedCompetitive citation presenceCompare branded and non-branded performance separately
AI referral trafficSessions arriving from recognised AI assistantsDownstream traffic from AI visibilityCompare engagement, conversion and pipeline quality with other channels

For B2B SaaS marketing, this creates a useful sequence: discovery first, citation second, visit third, pipeline last. The Clarity update improves visibility into the first two stages. It does not finish the attribution model.

Teams should also resist rewriting pages around every grounding query they see. Microsoft’s documentation makes clear that these queries can differ from the user’s original wording. For Google specifically, the practical guidance still centres on useful, accessible and well-structured content rather than creating pages for every possible query variation. Our guide on how to optimize content for AI search engines is a better starting point than chasing individual query strings.

What the dashboard still cannot tell you

The new segmentation adds useful context, but it is still important to define the limits of the data.

LimitationWhy it matters
It cannot reveal the exact user promptGrounding queries can differ from the original prompt, so branded queries should not be treated as proof that the buyer already knew the company.
Citation counts do not measure prominenceA citation shows that a page was referenced. It does not tell you whether the page was the first, most visible or most influential source in the answer.
Clarity is not a complete map of every AI platformSearch Engine Journal’s analysis recommends treating Clarity as a useful sample of AI citation visibility rather than assuming it represents every AI platform or interaction.

These limits do not make the dashboard less useful. They simply define what the metric is for. Marketers can use it to identify retrieval patterns, competitive citation gaps and pages that repeatedly contribute to AI answers, but they should not treat it as a universal market-share score for AI search.

What marketing teams should do next

Use branded vs non-branded queries as a segmentation layer, not as targets to inflate. A practical workflow is:

  1. Record the current branded vs non-branded mix. This gives you a baseline for how much citation activity is coming from brand-specific retrieval versus broader discovery.
  2. Map the cited pages behind each group. Look for which product, category, comparison, educational and proof pages are being retrieved most often.
  3. Prioritise commercially relevant non-branded gaps. If Share of Authority is weak around an important category, use case or problem, check whether you have a page that answers it clearly, contains current evidence and explains how your product fits.
  4. Audit branded information for consistency. Pricing, product names, features, integrations, security claims, customer proof and positioning should not conflict across pages.
  5. Connect visibility to downstream outcomes. Combine AI citation data with analytics and CRM data so you can see whether discovery translates into qualified visits, demos, opportunities and pipeline.

For non-branded queries, look for recurring themes rather than creating a new page because one grounding query appeared in the dashboard. For branded queries, focus on accuracy and consistency. If AI systems are performing brand-specific retrieval, your own website should provide the clearest and most current version of the facts.

The same principle applies to marketing measurement more broadly. Our guide to B2B marketing attribution explains why channel-level activity needs to connect to what happens after the form fill. Citation reporting should become another layer in that system, not a replacement for it.

OneMetrik Takeaway

Microsoft Clarity’s branded and non-branded query update is a relatively small reporting change with a useful strategic effect. It makes it harder to hide brand-specific retrieval and broader category discovery inside the same AI citation number.

At OneMetrik, we would use the split as a diagnostic. If branded visibility is strong and non-branded visibility is weak, the answer is not simply to “get more citations.” The better question is whether the website contains enough category expertise, product detail, evidence and proof for AI systems to understand where the company belongs before they perform a brand-specific lookup. If non-branded visibility is growing, the next question is whether that discovery creates meaningful visits, conversions and pipeline.

The metric that matters is not whether AI knows your name. It is whether AI can find you before the buyer names you.

Discover more from OneMetrik

Subscribe now to keep reading and get access to the full archive.

Continue reading