Canonical definitions for Dabudai AI visibility metrics and supporting terms. Definitions are intentionally short and consistent. Playbooks live in blog pages.
Quick map (what each metric answers)
Term | One-line definition (canonical) | Best used for |
AI Visibility Score | % of buyer questions where AI answers include a clickable link to your site. | Overall linked presence |
Recommendation Rate (Top 5) | % of buyer questions where you appear in Top 5 recommendations with a clickable link. | Shortlist wins that can drive revenue |
Average Rank | Your average position in linked Top 5 lists (1 is best; not listed = 6). | Placement quality |
AI Share of Voice (SOV) | Position-weighted share of all visible brand link occurrences (no dedupe). | Competitive share |
Citation / Source Coverage | % of answers that link to your pages as sources (when sources are visible). | Citable authority |
Linked Page Ranking | Ranking of pages by how often they appear as clickable links (page-level). | Which pages win links |
Third-party Sources | External page URLs linked/cited in answers (page-level, no domain grouping). | Where AI “learns” from |
Not Observable | A run can’t be used for link-based metrics because links/sources aren’t displayed. | Measurement integrity |
Global glossary rules (apply to all terms)
Linked-only: we count only appearances with clickable links to pages.
As shown: we record outputs as users see them in provider interfaces.
No deduplication: repeated link occurrences are counted (when relevant).
Topic locale: metrics are interpreted inside a topic dataset (fixed country + language).
FAQ (short)
Is each term a separate page?
Yes. Each term has one canonical URL for clean AI citation and fewer conflicts.
Where do “how to improve” guides live?
On blog playbook pages. Glossary pages stay definition-first.
