{"name":"Miragium AI visibility scorecard model","url":"https://www.miragium.com/assets/ai-visibility-scorecard-model","updatedAt":"2026-06-16","license":"Use for public AI visibility audits, GEO reports, prompt-testing scorecards, and citation-quality reviews. Treat scores as directional audit signals, not ranking guarantees.","rowCount":8,"rows":[{"metricKey":"prompt-coverage","label":"Prompt coverage","description":"How many high-intent prompts were tested for the topic.","inputField":"promptCount","formula":"min(promptCount / 20, 1)","weight":0.1,"goodThreshold":"20 or more prompts","warningThreshold":"10 to 19 prompts","actionWhenWeak":"Expand the prompt set across selection, comparison, source, and problem-aware queries.","aiRetrievalUse":"Use when an AI agent needs to judge whether the audit sample is large enough to trust.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"mention-rate","label":"Brand mention rate","description":"Share of tested prompts where the target brand or entity is mentioned.","inputField":"brandMentionCount / promptCount","formula":"brandMentionCount / promptCount","weight":0.2,"goodThreshold":"60% or higher","warningThreshold":"30% to 59%","actionWhenWeak":"Improve entity clarity, answer blocks, comparison pages, and public brand mentions around the topic.","aiRetrievalUse":"Use when diagnosing whether the entity is visible in AI answers at all.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"citation-rate","label":"Citation rate","description":"Share of tested prompts where the answer cites a source connected to the target brand or topic.","inputField":"citationCount / promptCount","formula":"citationCount / promptCount","weight":0.2,"goodThreshold":"45% or higher","warningThreshold":"20% to 44%","actionWhenWeak":"Create or improve citation-ready pages, markdown mirrors, datasets, and source lists for the tested prompt cluster.","aiRetrievalUse":"Use when a brand is mentioned but not cited with a source URL.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"source-accuracy","label":"Correct source rate","description":"Share of cited prompts that use the expected source URL or a closely related canonical URL.","inputField":"correctSourceCount / max(citationCount, 1)","formula":"correctSourceCount / max(citationCount, 1)","weight":0.15,"goodThreshold":"70% or higher","warningThreshold":"40% to 69%","actionWhenWeak":"Clarify canonical pages, citation graph links, dataset pages, and internal anchors for the expected source.","aiRetrievalUse":"Use when AI systems cite the wrong page or a weaker source for a claim.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"answer-accuracy","label":"Answer accuracy rate","description":"Share of tested prompts where the AI answer avoids unsupported claims and describes the product or topic correctly.","inputField":"accurateAnswerCount / promptCount","formula":"accurateAnswerCount / promptCount","weight":0.15,"goodThreshold":"80% or higher","warningThreshold":"60% to 79%","actionWhenWeak":"Add answer blocks, FAQs, claim boundaries, and proof rows that correct the common hallucination.","aiRetrievalUse":"Use when the answer mentions the brand but gets facts, positioning, or claims wrong.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"competitor-pressure","label":"Competitor pressure","description":"How often competitors are mentioned across the tested prompt set.","inputField":"competitorMentionCount / promptCount","formula":"1 - min(competitorMentionCount / promptCount, 1)","weight":0.08,"goodThreshold":"30% or lower competitor mention rate","warningThreshold":"31% to 60% competitor mention rate","actionWhenWeak":"Create comparison pages, alternative pages, and public proof assets that answer why the target entity is relevant.","aiRetrievalUse":"Use when competitors dominate answers even when the target brand has relevant pages.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"source-coverage","label":"Source coverage","description":"Number of public source types available for the prompt cluster, such as canonical pages, markdown mirrors, datasets, FAQs, and statistics.","inputField":"sourceTypeCount","formula":"min(sourceTypeCount / 5, 1)","weight":0.07,"goodThreshold":"5 or more source types","warningThreshold":"3 to 4 source types","actionWhenWeak":"Add public datasets, markdown mirrors, source lists, FAQ schema, or statistics pages to support the cluster.","aiRetrievalUse":"Use when the topic has too few citation surfaces for agents and search systems to retrieve.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"},{"metricKey":"freshness","label":"Freshness","description":"How recently the best source page or dataset was updated.","inputField":"daysSinceUpdate","formula":"daysSinceUpdate <= 30 ? 1 : daysSinceUpdate <= 90 ? 0.65 : 0.25","weight":0.05,"goodThreshold":"30 days or newer","warningThreshold":"31 to 90 days old","actionWhenWeak":"Refresh the source page, dataset updatedAt field, examples, screenshots, and cited statistics.","aiRetrievalUse":"Use when stale pages are weaker candidates for fast-moving AI search topics.","canonicalUrl":"https://www.miragium.com/assets/ai-visibility-scorecard-model","markdownUrl":"https://www.miragium.com/llms/ai-visibility-scorecard-model.md"}]}