A citation fingerprint is the distinctive blend of page types AI engines rely on when they cite sources for questions in one industry. It is a per-vertical pattern rather than a per-engine one: it describes what kind of page wins the citation — a listicle, an article, a homepage, a product page, a program page — not which engine does the citing.
The term comes from a July 2026 study by Brandon Kidd of DeltaV Digital, which analysed 25,337 citations across five AI engines and found the dominant cited page type varied sharply by category: listicles took about 61% of citations in B2B technology services, homepages about 55% for a local medical-aesthetics brand, and program pages about 53% in higher education (single-vendor data, eight brands, one per vertical — directional, not an audited benchmark). A related cut, own-domain citation share, ranged from about 74.7% in higher education to 0.0% in B2B technology services, showing that some verticals reward owned pages while others force earned placement.
For AI visibility, the citation fingerprint matters because it means the "right" content format is a property of your industry, not a universal rule. Mapping which page types get cited for your category's questions — and whether your own domain is ever among them — tells you what to publish and whether to pursue owned or earned coverage, which you then track as citation coverage and share of voice over time.