Most B2B domains receive zero citations from AI search engines. Machine Relations Index data tracking 7,096 enterprise domains across six engines — Perplexity, ChatGPT, Gemini, Claude, Google AI Mode, and Google AI Overviews — shows that citation authority follows an extreme power law. A fraction of domains capture nearly all citations. The rest are structurally invisible, and the reasons are measurable.
How Concentrated Is AI Citation Authority? #
Among 7,096 enterprise domains tracked by the Machine Relations Index across 28,625 source events, the distribution is stark. The top-ranked market database domain — G2 — recorded 210 citations across all six engines, 42 distinct queries, and 10 verticals in the most recent 30-day window. Crunchbase recorded 198 citations across 35 queries. Gartner, operating as analyst research, accumulated 232 citations across 70 queries.
These are the 99th-percentile domains. The median enterprise domain in the same dataset records zero or single-digit citations — often from only one engine and for a narrow query set. The gap is not incremental. It is orders of magnitude.
This pattern mirrors what Smart Money Media found when measuring 30 companies across 5,400 AI engine responses: funded B2B startups were cited only 10% of the time, compared with 34% for established incumbents. The gap was 3.4x overall and far worse in specific categories — legal tech AI startups were cited just 2% of the time.
Five Structural Factors That Separate Cited From Invisible #
The Machine Relations Index scores source authority using five measurable components. These are not arbitrary — they reflect the structural signals AI engines use to decide which sources to surface.
| Factor | What It Measures | Elite Threshold | Why It Matters |
|---|---|---|---|
| Engine breadth | How many of the six engines cite the domain | 6/6 (score 40) | A source cited by only one engine has no consensus authority |
| Query diversity | Distinct query intents the domain appears for | 35+ queries (score 14+) | Single-query sources are niche; multi-query sources are reference material |
| Vertical spread | How many industry verticals the domain covers | 9-10 verticals (score 15) | Engines prefer sources that demonstrate cross-domain relevance |
| Position quality | Average citation position across appearances | Top 5 (score 3+) | First-cited sources compound because engines reuse them |
| Temporal consistency | How many days the domain appeared in the measurement window | 23+ of 28 days (score 8+) | Engines drop sources that appear intermittently |
A domain scoring zero on engine breadth — cited by no engines — cannot score on any other factor. This is the structural trap: most B2B domains never reach the minimum threshold on even the first factor because they lack the third-party editorial coverage, structured data, and web-wide entity signals that engines use to discover sources in the first place.
Smart Money Media's research identified three root causes for the startup visibility gap: editorial coverage disparity (incumbents have deeper tier-1 coverage and denser Wikipedia/Wikidata footprints), the failure of funding PR to translate into AI citation authority, and entity graph gaps where startups lack the structured identity signals engines rely on for source selection.
Ghost Citations Make the Problem Worse Than It Appears #
Even among domains that do receive citations, visibility is overstated. Research from Ivristech analyzing 3,981 domain appearances across four AI platforms found that 61.7% of AI citations are "ghost citations" — the domain receives a source link, but the brand name never appears in the answer text users actually read.
Only 13.2% of domain appearances resulted in both a citation link and a brand name mention simultaneously. Citation metrics alone overstate actual brand visibility by roughly 4–6x.
The engine-level variation is significant. ChatGPT cites sources 87% of the time but mentions brand names in only 20.7% of appearances. Gemini mentions brands at 83.7% but provides source links for only 21.4%. This means a brand that appears in ChatGPT results may receive traffic from links without any brand recognition effect, while a brand that appears in Gemini results may get brand exposure without direct referral traffic.
For the zero-citation majority, this compounds the problem. Even if a B2B domain manages to break through to a single engine, the odds of achieving meaningful brand visibility — not just a buried footnote link — are low.
Why Funding and Domain Authority Do Not Solve This #
The assumption that well-funded companies with strong traditional SEO profiles will naturally appear in AI engines is wrong. Smart Money Media's controlled study tested this directly: among 19 funded startups with active investor PR, the average AI citation rate was 10%. Category-specific results were worse — legal tech AI at 2%, cybersecurity at 7%.
The disconnect exists because AI engines do not use the same ranking signals as traditional search. Machine Relations Index data shows that citation authority is built through third-party editorial coverage across multiple publications, consistent structured data across entity graphs, cross-vertical query relevance, and temporal persistence — appearing repeatedly over weeks, not just during a launch cycle.
Domain authority, backlink counts, and organic keyword rankings — the metrics B2B marketing teams optimize for — have limited predictive power for AI engine citation selection. A domain with a DR of 70 and no third-party editorial coverage in the entity graph will still score zero on engine breadth.
What the Machine Relations Index Reveals About the Citation Threshold #
The MRI methodology measures citation authority at the domain level across six engines using a composite score with a maximum of 100. The current dataset spans 7,096 enterprise domains.
The Elite tier — domains with consensus scores above 77 — represents fewer than 15 domains in the full dataset. These are sources like G2 (MRI 80.8, rank 1 among market databases), Crunchbase (MRI 80.3, rank 2), Gartner (MRI 77.2, rank 2 among analyst firms), and Deloitte (MRI 78.4, rank 1 among analyst firms). Each is cited by all six engines, across 28–70 distinct queries, in 9–10 verticals, with temporal consistency of 22–28 days out of 28.
The vast majority of tracked domains — those outside the top 50 — score below 30 on the composite index. Many score zero: no citations from any engine in the measurement window. These are not low-quality domains. They include funded SaaS companies, established B2B platforms, and enterprise software vendors with real customers and real products. They are structurally invisible because they lack the specific combination of third-party coverage, entity graph presence, cross-vertical relevance, and temporal citation consistency that engines use for source selection.
This is the zero-citation problem. It is not a content quality issue. It is a structural authority issue that operates on different signals than traditional search ranking.
FAQ #
What percentage of B2B domains get zero AI engine citations? #
Based on Machine Relations Index data tracking 7,096 enterprise domains, the large majority record zero or single-digit citations from AI engines in any given 30-day window. The top 10 domains by citation volume account for a disproportionate share of all citations across 28,625 source events. Funded startups specifically were cited only 10% of the time in controlled testing across 5,400 AI engine responses.
Does high domain authority help with AI engine citations? #
Not directly. AI engines use different signals than traditional search rankings. Machine Relations Index components — engine breadth, query diversity, vertical spread, position quality, and temporal consistency — are driven by third-party editorial coverage, entity graph presence, and cross-domain relevance rather than backlink profiles or keyword rankings.
What is a ghost citation in AI search? #
A ghost citation occurs when an AI engine links to a domain as a source but never mentions the brand name in the answer text. Research shows 61.7% of AI citations are ghost citations, and only 13.2% of domain appearances result in both a citation and a brand mention. This means standard citation tracking overstates actual brand visibility by 4–6x.
How can B2B brands break out of zero-citation invisibility? #
The structural factors are measurable: engine breadth (being cited by multiple engines), query diversity (appearing for many query intents), vertical spread (relevance across industries), position quality (appearing early in citations), and temporal consistency (persistent appearance over time). Each requires building third-party editorial coverage, structured entity data, and cross-domain relevance — not just publishing more content on owned domains. The Machine Relations framework provides a methodology for measuring and improving these signals.
Last updated: June 15, 2026. Data from Machine Relations Index v1.1 (6-engine methodology) tracking 7,096 enterprise domains across 28,625 source events.
Additional source context #
- Famous but Invisible: Why AI Talks About Your Brand and Never Links to You | by Sebastian Mueller | The Citation Lab | Jun, 2026 | Medium Sign up Get app Sign up ## The Citation Lab What gets you cited by AI — and what doesn’t. (Famous but Invisible: Why AI Talks About Your Brand and Never Links to You | by Sebastian Mueller | The Citation Lab | J, 2026).
- 62% of AI Citations Don’t Mention Your Brand: The Ghost Citation Problem – Marketing Agent Blog ## 62% of AI Citations Don’t Mention Your Brand: The Ghost... (62% of AI Citations Don’t Mention Your Brand: The Ghost Citation Problem – Marketing Agent Blog (marketingagent.blog), 2026).
- Why 90% of What AI Says About Your Brand Isn't From You [Report] ‹ Go Back ## For Exclusive Trends, Research & Data Gain access to exclusive research, training, trends and support from the best marketers in the world. (Why 90% of What AI Says About Your Brand Isn't From You [Report] (foundationinc.co), 2026).
- And in more than half of cases, a citation doesn’t instantly lead to a brand mention. (Why 62% of AI citations don’t lead to brand mentions [Study] - mGrowTech (mgrowtech.com), 2026).
- Why doesn't AI cite my B2B company when buyers ask for recommendations? | PitchKitchen provides external context for why most B2B brands get zero AI engine citations.
- Why AI Engines Cite Some Brands and Ignore Others (2026 Research) | Magna provides external context for why most B2B brands get zero AI engine citations.
- LLM Ghost Citations: Why Your Content Is Working and Your Brand Isn't provides external context for why most B2B brands get zero AI engine citations.
- The ghost citation problem - by Kevin Indig - Growth Memo provides external context for why most B2B brands get zero AI engine citations.
- Why Your Brand Isn't Cited by ChatGPT: 10 Common Causes (and Fixes) | LumenGEO provides external context for why most B2B brands get zero AI engine citations.