Microsoft Clarity AI Visibility: measure citations with context

AI citations are a new visibility signal. Microsoft Clarity lets a business inspect them alongside visits and sales, within its supported scope.

What AI Visibility measures

Microsoft has made Citations in Clarity generally available. It reports when pages from a domain are cited in supported AI answers, with page citations, grounding queries, a comparative citation share and AI referral traffic. The supported experiences include Microsoft Copilot and partner environments. This is a view of those sources, not a census of everything said by every AI model.

Citation does not equal a customer

One answer may cite pages from a domain more than once. Citation counts therefore do not represent people, sessions or sales. Clarity separately reports the proportion of sessions referred by AI assistants. Compare that with engagement, enquiries and revenue. If a guide gets cited but has no useful route to an offer, improving the page journey may matter more than adding another article.

Branded versus generic queries

Clarity now separates grounding queries mentioning your brand from broader category questions. Being cited when your name is already known is different from appearing during supplier discovery. The dashboard also breaks out Share of Authority by query type. That metric is calculated within a set of citations and queries; it must not be presented as a share of the entire AI market.

Find useful editorial gaps

Compare highly cited URLs with your service pages and the buying journey. Do pages answer a specific question, give current evidence and suggest a sensible next step? Look for topics where competing domains are cited while yours is absent. Improving a page with an information gap can be more valuable than publishing near-duplicate posts about the same question.

Set up a sensible review

Create a Clarity project and verify domain ownership where required, then review AI Visibility → Citations. Record a monthly baseline by page and by branded versus non-branded queries. After edits, inspect citation trends and verify visits and leads in analytics. Clarity and Google Search Console measure different surfaces: do not add their figures together as one visibility score.

Known brand, weak category presence

A firm may be cited frequently when the query already contains its name but rarely when someone asks about the broader problem it solves. Those are different positions in the buying journey. Separating branded and generic grounding queries prevents an inflated view of category visibility. First inspect pages explaining use cases, evidence and limits; only then decide whether a new article is needed.

Keep signals separate

An AI bot fetching a page is not a person and does not prove a citation. A citation need not generate a click; a click need not become an enquiry. Bot activity, Citations, referral sessions and CRM outcomes cover different stages. Comparing them side by side can expose the bottleneck, but summing them into one score would obscure it.

An editorial routine

Each month choose a few non-branded questions relevant to the buying process. Compare cited pages with the pages that should cover those questions. Improve the specificity and freshness of the answer and add a useful route to the service. Log the edit and follow the trend for several weeks. If citations rise without leads, review intent fit and the next step on the page.

FAQ

Does Clarity cover every ChatGPT and Google answer?

No. Citations covers supported AI experiences; it is not a complete monitor of every platform.

Is Share of Authority total market share?

No. It compares citations within the query set covered by the dashboard.

Sources

Information checked: 25 September 2026. Feature access may depend on the account.