Ghost Citations in AI Search Are Erasing Brand Visibility You Think You Have
The Development
A study conducted by SEMrush in partnership with analyst Kevin Indig, published June 9, analyzed 3,981 domain appearances across ChatGPT, Gemini, Google AI Overviews, and Google AI Mode. The central finding: 62% of AI citations do not result in brand mentions. These so-called ghost citations — where an AI engine references a source URL without surfacing the brand name in the generated answer — mean that a significant majority of AI-driven referral credit accrues to domains without translating into brand recognition or recall for the user. Separately, AIMZER launched the AI Visibility Framework on the same date, a proprietary methodology tracking seven AI-specific metrics across ChatGPT, Gemini, Claude, Perplexity, Copilot, Grok, and Meta AI, including Mention Frequency, Citation Rate, Entity Accuracy, and Competitive AI Share of Voice.
Our Take
The 62% ghost citation figure exposes a fundamental flaw in how most brands are currently evaluating their AI search presence: citation counts are being treated as a proxy for visibility when they are not. A citation buried in an LLM's source footnote, never surfaced to the user, delivers zero brand equity. The practical consequence is that share-of-voice reporting built on citation tracking alone systematically overstates AI-driven brand exposure. What actually governs consumer discovery in AI search is whether the brand name appears in the generated answer — and most measurement stacks are not tracking that distinction. The AIMZER framework and the SEMrush data together signal that AI visibility is fragmenting into at least three distinct metrics: citation presence, brand mention frequency, and recommendation consistency, each requiring separate optimization strategies.
What Changed
Brands can now measure the structural gap between AI citation volume and actual brand mention frequency across major AI platforms. This disaggregation — citation rate versus mention rate versus recommendation consistency — creates a new performance surface that legacy search analytics tools do not track.
Marketing Impact
Brand and media measurement teams are most directly affected. Current AI visibility dashboards built around referral traffic and citation counts are producing materially misleading performance data, underreporting visibility gaps and overreporting brand exposure in AI-generated search environments.
Competitive Implication
Brands that instrument the citation-to-mention gap now gain an accurate map of where AI engines are erasing their name from consumer-facing answers. Brands still reporting only citation volume are optimizing against a metric that has no direct relationship to how AI search is actually shaping purchase consideration.
Strategic Outlook
Expect measurement vendors including Semrush, Conductor, and BrightEdge to productize AI mention frequency and recommendation consistency as standard reporting dimensions by Q3 2026. The category AIMZER is calling AEO will consolidate into the broader SEO toolchain rather than remaining a standalone discipline.
The Exploit
Action Item
Heads of brand measurement should run a citation-versus-mention audit against their top ten category keywords across ChatGPT and Google AI Mode this week, using the gap as the forcing function to update their AI visibility KPI framework before Q3 budget reviews.