Open a question
Choose Explore, then Questions. Filter to the service, language or market you want to inspect. Open a question to read its recorded answer and measurement context. Separate unbranded buyer questions from branded or comparison questions.
Read the answer and its evidence
Check the wording, AI service, locale, date and method before interpreting the answer. Named means the brand appears in the recorded response; it does not automatically mean a recommendation. Not named is a valid observation. No usable answer means the response was unavailable, not proof of no visibility. Read the attached sources to check what supports the answer.
Sources and source gaps
Sources groups cited pages and domains. A cited third-party page is not your owned page. Open the question references to see the answer context. Source gap analysis compares recorded source coverage and links to relevant evidence or work where available. It does not prove that publishing on a domain will cause an AI recommendation.
Use the selected scope
Filters apply to the current view. Clear them if no rows match. A CSV records the exported table scope; keep the date, service and question identity when sharing it. Do not combine historical API and browser observations as one comparable score.
Frequently asked questions
Does named mean recommended?
No. A mention and a recommendation are separate observations.
Is no answer the same as not named?
No. No usable answer is missing evidence, while not named describes a recorded answer.
Can a source gap predict an improvement?
No. It identifies an evidence gap, not a guaranteed effect of publishing.