Afternoon BriefAI Search & Discovery

AI Knows Your Brand. It Just Won't Say Your Name.

AI platforms recognize 96% of brands but name fewer than 11% in buying-stage answers. Here is how to diagnose your ghost citation rate and cross the mention threshold.

Christian Lehman
Christian LehmanAug 4, 2026

AI platforms can describe your company, your product, and your market position with 96 percent accuracy. But when a buyer asks which brand to consider, 89 percent of those same brands never get named. That gap has a name now — ghost citations — and it is the single largest leak in your AI visibility funnel. With 68 percent of Google searches ending without a click and AI Overviews appearing on nearly half of all queries, whether your brand gets named in the answer is now the game. Here is what the latest data shows and what to do about it.

62 Percent of AI Citations Never Name Your Brand

A Semrush study with Kevin Indig analyzed 3,981 domain appearances across ChatGPT, Gemini, Google AI Overviews, and Google AI Mode. The finding: 61.7 percent of all AI appearances were ghost citations — the engine linked to your page as a source but never said your name in the answer text. Only 13.2 percent of appearances included both a citation link and a brand mention.

This means you can win the citation and still lose the customer. As Seer Interactive put it: you are probably funding your competitors' recommendations without realizing it. The buyer reads the answer, gets the information your content provided, and moves on without ever knowing it came from you.

Why AI Recognition Does Not Equal AI Recommendation

Victorious tested 175 brands across eight AI platforms in their Q2 2026 Quarterly Search Report. When asked directly about a brand, AI platforms described it accurately 96 percent of the time. They knew the product, the market, the positioning.

But when the researchers switched to buying-stage research prompts — the questions a real prospect asks when comparing solutions — 89 percent of those brands never appeared in the answer. The AI knew them. It just did not recommend them.

Even more telling: 99.99 percent of the 49,391 citations in buying-stage answers pointed to third-party websites. Only four of 150 brands earned a citation to their own domain during the consideration phase. The full Victorious Q2 report breaks down the visibility benchmarks at each threshold.

The 2,000-Mention Threshold

The Victorious data revealed a hard floor. Brands with fewer than 2,000 indexed web pages mentioning them were named in AI answers just 3 percent of the time. Above that threshold, mention rates climbed steeply.

The strongest correlations came from two signals: referring domains (0.49 correlation with AI mentions) and third-party web mentions (0.45 correlation). Link quality alone did not move the needle — some brands with high-authority backlink profiles still rarely appeared. Volume of independent references mattered more than the authority of any single one.

AirOps' State of AEO 2026 report confirmed the pattern: 85 percent of brand mentions in AI answers come from third-party sources, not the brand's own website.

Every Engine Behaves Differently

You cannot treat AI visibility as one metric. The Semrush study found that each engine has a distinct personality when it comes to citations versus mentions:

  • Gemini mentions brands in 83.7 percent of appearances but only generates a citation link 21.4 percent of the time. It talks about brands from memory.
  • ChatGPT does the opposite: 87 percent citation rate, but only 20.7 percent mention rate. It footnotes your page without naming you.
  • Google AI Overviews sit in the middle, leaning toward citations.
  • Google AI Mode mentions brands at nearly twice the rate of ChatGPT, but still behaves more like a research footnote than a recommendation.

The engines disagreed on whether to name a specific brand 22 percent of the time across the same prompts. A brand named by Gemini was often ghost-cited by ChatGPT for the same query.

How to Diagnose and Fix Your Ghost Citation Rate

This is not a branding problem. It is a structural one. Here is the diagnostic:

Step 1: Measure citation and mention separately. If you are only tracking whether your domain appears as a source, you are missing whether your brand actually shows up in the answer text. Run the same prompt set across ChatGPT, Gemini, and Google AI Mode and log both signals independently.

Step 2: Audit your third-party mention volume. Search "your brand name" -site:yourdomain.com and count indexed pages. If you are below 2,000, that is the bottleneck — not your on-site content.

Step 3: Create content that triggers naming, not just citation. The Semrush data showed that comparative content produces 2.4 times more brand mentions than informational content. Short, conversational queries generated 30 to 50 times more brand mentions than long prompts. Structure your content around direct comparisons and clear brand attribution in the first sentence of every section.

Step 4: Build third-party presence where AI actually looks. At the buying stage, AI engines pull from review sites, comparison pages, industry publications, and community discussions — not your blog. Earned media, directory listings, and analyst coverage feed the recommendation layer. Your owned content earns citations at the awareness stage; third-party presence earns mentions at the decision stage.

Step 5: Refresh on a 90-day cycle. Pages updated within the last three months are three times more likely to be cited, according to AirOps. A Pillarbase study of 15.7 million AI Mode citations found that 47.7 percent of citations are scroll-to-text highlights — AI cites specific passages, not pages. If those passages go stale, the citation drops regardless of how good the surrounding content is.

Where Machine Relations Fits

This is the gap that Machine Relations was built to close. Traditional SEO gets your pages indexed. PR gets your name in publications. But neither discipline specifically targets the mechanism that makes AI engines name your brand in a buying-stage answer — which is consistent, frequent, third-party presence across the sources each engine trusts.

The ghost citation problem is a measurement gap as much as a visibility gap. If you are only tracking citations, your AI visibility dashboard looks healthy while your brand stays invisible to buyers. Measure mentions. Build third-party volume. Structure for naming, not just linking.

FAQ

A ghost citation occurs when an AI search engine uses your page as a source but never names your brand in the answer text the user reads. According to Semrush research, 62 percent of AI citations are ghost citations — the brand gets attribution in a footnote link but remains invisible in the actual response.

How many third-party mentions does a brand need to appear in AI answers?

Victorious found that brands with fewer than 2,000 indexed web pages mentioning them were named in AI answers just 3 percent of the time. The mention rate increases significantly above that threshold, driven by volume of independent references rather than the authority of any single source.

Do different AI engines cite and mention brands the same way?

No. Gemini mentions brands by name 83.7 percent of the time but rarely provides a source link. ChatGPT provides source links 87 percent of the time but names the brand in only 20.7 percent of answers. A brand can be visible on one platform and ghost-cited on another, which is why measurement needs to be per-engine.