Mention and citation rates: denominators for AI-answer reporting

Answer-engine reporting needs explicit denominators. Compare mention and citation rates across a fixed prompt roster, with denominators and capture dates shown. A single composite visibility metric is withheld until one versioned method reconciles product views and rollups.

Why the denominator matters

A raw mention count can change when the prompt roster, provider behavior, or collection coverage changes. A rate is interpretable only with its numerator, denominator, prompt set, surface, and capture window.

What to record

A fixed prompt roster per project, with language, market, and capture settings documented.

Separate collection by supported answer surface so like-for-like prompts can be compared.

Mentions and citations counted separately, with failed or unconfigured probes visible and excluded from relevant denominators.

Competitor comparisons when snapshots exist, with the comparison roster and denominator documented.

Report rates with their evidence and limitations rather than collapsing them into one headline number. Source lists can guide review; they do not prove referral traffic or ranking impact.