Skip to main content

​
Monitoring and data

It represents an answer captured for a specific prompt, model, country, persona, and time. Model output can vary between generations, so one answer is an observation rather than a universal result.
Yes. A prompt can inherit the workspace default or use Daily, Weekly, Monthly, or No monitoring. You can also pause or resume individual prompts from Workspace Settings > Prompts.
A boost generates multiple answers for the same prompt and model in one run. You can set 2 to 20 answers. Boosts increase credit use, and the workspace limits how many prompts can be boosted at once.
The answer-monitoring pipeline does not require a crawler or tracking tag on your site. Other workflows may fetch public pages you submit, including onboarding analysis, Site Diagnostics, and source analysis.

​
Metrics

They use different units and denominators. Mention Rate is answer-level and considers answers with a SELF or DIRECT brand mention. Dashboard Coverage is prompt-level and considers all distinct answered prompts in scope. See Core concepts.
The REST performance endpoint calculates answers with a SELF mention divided by all answers. The dashboard card calculates distinct answered prompts with a SELF mention divided by all distinct answered prompts. Use the definition for the surface you are reporting.
Mention Rate, Citation Rate, and Share of Voice use SELF and DIRECT relationships. Reclassifying a detected brand changes which records enter those calculations for the selected scope.

​
Topics and prompts

Qwairy keeps the prompts and unassigns them from the deleted topic. It also removes the topic from saved taxonomy references and cleans related topic projections. Assign another topic if you need those prompts in topic-level views.
Yes. A prompt can have multiple tags but at most one topic.

​
Credits and access

Answer generation, Perception analysis, completed Site Diagnostics page audits, Content Studio article or brief creation and duplication, and Pitch Audit answer generation consume credits. Creating or importing prompt definitions is free. See Providers and credits.
For each prompt, add up the costs of the models available in each of the prompt’s countries, then multiply the total by the number of personas and by the boost. The Monitoring page shows an estimate before you save the configuration.
The team Owner can buy a one-time credit pack under Team Management > Billing. Credit packs require an active paid subscription. Sub-teams receive credits allocated by their parent team.
No. Fact Check is a beta feature and does not debit customer credits.
Use Workspace Settings > Shared Links for shareable views and Workspace Settings > Exports for generated files. Team integrations include the REST API, MCP Server, and Looker Studio where enabled for the account.

​
Troubleshooting

Check the current workspace, date range, saved list, and filters. Confirm that answers exist for the selected prompts and models. Some insight pages appear only when the corresponding answer metadata was detected.
Google Search, including AI Overviews and AI Mode, crawls with Googlebot, which Crawler Analytics excludes because its requests cannot be attributed to AI features. The Google identities that are counted serve narrower uses, so low volumes are expected. Read Why does Google rarely appear in Crawler Analytics? for the full explanation and the data to use instead.
Use Monitor > Competitor Mentions to change a brand relationship. Use Monitor > Source Explorer to inspect source ownership and type.