Semrush, an Adobe company, this month published the AI Visibility Index 2026, a study analyzing 126 million United States AI search prompts collected between January and April 2026. The research, which examined how ChatGPT, Google Gemini, Google AI Mode and Google AI Overviews describe and cite more than 1,200 brands across 22 industry verticals, found that only 36 of those brands appeared in the top 100 most-mentioned list on every platform in every month of the study window. Semrush calls this group the Universal 36.

The report arrived through an embargoed briefing distributed by Victoria Shigalova, a public relations manager at ZECOMMS Agency, on June 24, 2026, ahead of a planned June 25 release. Semrush's own press materials describe the prompt database as cumulative, growing from 89 million deduplicated prompts in September 2025 to 126 million by April 2026, with new prompts added whenever a keyword begins triggering an AI-generated answer.

The Universal 36 includes platforms that already function as default destinations for large swaths of digital life: YouTube, Google, Reddit, Facebook, Amazon, Instagram, Apple, eBay, Walmart, Etsy, Yelp, TikTok, Microsoft, Netflix, Disney, Home Depot, LinkedIn, IMDb, Ford, Spotify, Indeed, Adobe, FDA.gov, Pinterest, Samsung, Toyota, NFL.com, MLB.com, Chevrolet, Marvel, Honda, Prime Video, NBA.com, Best Buy, USDA.gov and Nintendo. According to the report, mention volume trails off sharply within that group, falling from YouTube's 47.9 million mentions at the top to Nintendo's 1.5 million at the bottom of the list.

What separates a mention from a citation

The central distinction running through the study is the difference between two metrics that Semrush treats as structurally separate: a mention, which is when an AI system names a brand inside its answer, and a citation, which is when the system quotes from that brand's own website as a source. According to the report, a brand can be mentioned constantly and cited almost nowhere, or cited constantly while almost never being the brand the AI is actually discussing. Wikipedia is offered as the clearest example of the latter pattern, heavily quoted as a source while rarely the subject of the conversation itself.

The gap between the two metrics varies sharply by platform. Across the four AI systems Semrush measured, the overlap between the brands that get mentioned and the brands that get cited as sources, what the report calls mention-source overlap, ran as high as 64% on Google AI Overviews and as low as 30% on Gemini, a spread of 34 percentage points. On AI Overviews, most of the brands AI talks about are also brands being quoted, while on Gemini they're largely separate groups.

This separation matters because the two metrics are earned through different mechanisms. Mentions come from brand authority and category fit. Citations come from content depth and what the report terms a Source Surplus, defined as the gap between how often a website is cited and how often that same brand is mentioned by name. The highest Source Surplus brands identified in the dataset include Medical News Today at a 4.6x citation-to-mention ratio, Wikipedia at 4.3x, Healthline at 4.2x and WebMD at 4.0x. Each of these sites functions, in the report's framing, as infrastructure that AI systems use to construct answers about other brands entirely.

How the four platforms diverge

Semrush measured citation density, meaning the average number of sources an AI platform cites per response, and found wide variation. ChatGPT cited an average of 15.4 sources per response, the highest of the four platforms, drawing heavily on Reddit and Wikipedia. Google AI Mode followed at 11.4 sources, leaning toward social and local platforms such as Facebook, Instagram and Yelp. Google AI Overviews cited 9.2 sources on average, dominated by YouTube as the single largest source. Gemini cited just 3.3 sources per response, the lowest by a wide margin, concentrating its citations on entertainment and commerce sources including YouTube, Wikipedia and e-commerce platforms.

A structural reason underlies part of that gap. According to the report, the average AI Overviews prompt runs about 24.6 characters, against 57.8 to 58.1 characters for prompts directed at ChatGPT, Gemini and AI Mode. Semrush attributes the difference to where the prompts originate: AI Overviews queries come from Google Search keywords rather than conversational input, and shorter queries tend to produce shorter answers drawing from fewer sources.

Brand mention density follows a different pattern from citation density. Google AI Mode named the most brands per response at 5.2, followed by Gemini at 4.7, AI Overviews at 3.7 and ChatGPT at 3.3. Google AI Mode names the most brands per response, around 58% more than ChatGPT, which differs from the citation density pattern, where ChatGPT leads and Gemini trails.

Each platform also showed a distinct top-10 ranking when Semrush aggregated mentions from January through April 2026. On ChatGPT, Google and Reddit led, followed by a run of commerce and social platforms including Walmart, eBay, Amazon, Facebook, Instagram and Etsy. On Gemini, Google itself topped the list at 18.3 million mentions, consistent with Gemini's integration across Google's own product surfaces, with the remainder of the list leaning heavily toward commerce brands such as Amazon, eBay, Walmart and Etsy. On Google AI Mode, YouTube led at 6.4 million mentions, but Yelp's appearance at sixth place, with 3.6 million mentions, stood out as the only instance across any of the four platforms where a local-discovery brand reached that high a position. On Google AI Overviews, YouTube dominated at 35.2 million mentions, more than double the next brand, Reddit, at 15.2 million.

Only three brands, YouTube, Amazon and Facebook, appeared in the top 10 of every single platform. Beyond that intersection, the four rankings diverged considerably.

Patagonia and Shopify as case studies

Semrush singled out two brands as detailed case studies in the report: Patagonia and Shopify. Patagonia's AI Visibility score, a 0-100 metric Semrush uses to represent how prominently a brand appears across the four platforms combined, held at 79 or 80 in every month between January and April 2026. The brand scored 76 on ChatGPT, 78 on Gemini, 76 on AI Mode and 87 on AI Overviews, generating roughly 87,000 total mentions and 8,300 cited pages across the platforms.

According to the report, that consistency stems less from paid media or content volume than from a network of third-party outdoor-gear review sites. Six specialist sites, led by Reddit at 38,800 mentions, OutdoorGearLab at 18,500 and REI at 12,900, collectively generated more than 65,000 brand mentions for Patagonia, a figure Semrush says exceeds the combined output of the entire Tier 1 traditional media covered in the dataset.

Shopify presented a different pattern. The company's overall AI Visibility score held flat at 81 across the same four-month window, with Google AI Mode reaching 86 and AI Overviews reaching 84. Total mentions across platforms reached 147,000, against 42,100 cited pages, the deepest citation footprint among the case studies in the report. On Google AI Overviews specifically, Shopify was mentioned 45,098 times and cited 46,342 times in April 2026, a near-identical balance that the report describes as rare; most brands skew toward being either predominantly mentioned, as with Patagonia, or predominantly cited, as with Wikipedia.

Semrush attributes Shopify's balance to three layers operating in parallel: specialist software-review platforms such as G2, which ranked fifth in Shopify's ChatGPT citation mix with 1,124 mentions, and Capterra, which ranked eighth with 808; topic and sub-brand authority, with Shopify maintaining 222 individual documentation topics scoring above 90 on the AI Visibility scale and a distinct Shopify Plus sub-brand scoring 64 on its own; and a long tail of community sources including YouTube at 76,100 citations, Reddit at 44,500 and LinkedIn at 15,400.

Concentration varies sharply by industry

Semrush measured what it calls a Concentration Curve across the 22 verticals in the dataset, defined as the share of category top-10 mention volume held by the top three brands. The spread between categories reached 41.5 percentage points. News & Media was the most concentrated category at 82.9%, with Reddit alone holding 25.8 million mentions, more than twelve times the next-ranked news brand, Yahoo, at 2.1 million. Consumer Electronics followed at 76.9%, driven by Apple's 11.2 million mentions, five times the volume of second-place Samsung at 2.1 million.

Finance sat at the opposite end of the curve at 41.4%, the flattest category measured. Five brands clustered near the top: PayPal led with 775,695 mentions, followed by Chase at 650,610, Intuit at 633,626, Investopedia at 577,015 and NerdWallet at 549,119. Industrial followed closely at 42.2%, with Boeing, 3M, John Deere, Milwaukee Tool and DeWalt occupying the leading positions, though Semrush noted that the category carries lower absolute mention volume overall, which the report interprets as evidence that AI systems currently underweight business-to-business queries relative to consumer categories.

Apple's position within Consumer Electronics also illustrated a separate pattern the report calls sub-brand depth. According to Semrush, 18 distinct Apple sub-brands each registered at least 1% of the parent brand's total mentions, including Apple Music at 263,000 mentions, Apple TV at 260,000 and iPhone at 143,000. By comparison, the report states that Patagonia and Shopify each register zero sub-brands at that threshold, and Cleveland Clinic registers one, making Apple a structural outlier in how granularly AI systems track its product lines.

Marketers say they cannot measure what is happening

Several of the report's findings draw on a separate Semrush survey of 481 marketers regarding AI search behavior. According to that survey, 45% of respondents said they cannot properly measure their brand's visibility in AI-generated answers, and only 9% said they can measure all the metrics the report identifies as relevant: per-platform mentions, per-platform citations, mention-source overlap and cross-platform consistency.

The same survey found a sharp split tied to organizational structure. Eighty-one percent of marketers with fully integrated SEO and AI search execution reported more traffic or leads from AI platforms, compared with 36% among teams running completely separate workflows. Semrush frames this gap as the single largest factor identified in the research, larger than any platform-specific tactic covered elsewhere in the study.

The measurement gap that the survey describes is consistent with broader reporting on the AI search measurement landscape. Similarweb research published in November 2025 found that citation sets across AI platforms change by roughly 50% on a monthly basis, with only 11% overlap in citations between major platforms, a volatility pattern that helps explain why a high proportion of marketers report difficulty tracking visibility consistently over time. Separately, Adobe's own April 2026 survey of more than 500 marketers found that 98% lacked a clear, documented and confident roadmap for AI search optimization, with 74% reporting either no measurable strategy at all or no awareness of one within their organization.

The report frames AI visibility as cross-functional

Semrush's report proposes a framework it calls the four layers of brand visibility: discoverability, meaning whether AI systems can find and retrieve a brand's content, owned by SEO and content teams; clarity, meaning whether AI understands the brand correctly, owned by product marketing and brand teams; authority, meaning whether the brand appears qualified to be included in an answer, owned by public relations, partnerships and affiliate teams; and trust, meaning whether AI will confidently recommend or act on the brand, owned by social, community and customer service functions.

The report also identifies what it calls three AI visibility clusters into which most brands fall. The Commerce cluster, including Amazon, eBay, Walmart, Etsy and Apple, wins through transaction scale and product depth but carries heavy mentions with comparatively light citations, since AI talks about commerce brands without quoting directly from their pages. The Community cluster, including Reddit, Wikipedia, Quora and Fandom, wins through user-generated content at scale and earns both mentions and citations, but the report flags a vulnerability: Wikipedia's mentions reportedly declined by 50% between January and April 2026 even as its citation volume remained strong, which Semrush interprets as a sign that AI systems are increasingly able to synthesize information without naming the source brand. The Network cluster, including Patagonia, Shopify, Notion, LEGO, Cleveland Clinic and NerdWallet, wins through citation infrastructure built across third-party sources rather than owned content, leaving these brands more exposed to changes in how those third-party sites operate or rank.

The report also flags platform concentration as a specific risk category. Cleveland Clinic draws 86% of its total AI mentions from Google AI Overviews alone, according to the data, meaning any shift in how that single platform weights health sources could move the bulk of the brand's AI visibility. Yelp shows a similar pattern, with 704,000 of its mentions, nearly half its total, concentrated on Google AI Mode. Home Depot was offered as the inverse example, with mentions spread within a 3.4 percentage point band across all four platforms, a distribution the report calls the most even in the dataset.

Context for the broader AI search measurement landscape

The Semrush release follows a year in which AI search measurement has emerged as a recurring point of friction for marketers. Semrush itself documented in October 2025 that it nearly tripled its own AI share of voice from 13% to 32% in a single month after discovering that ChatGPT recommended every competitor in its category but never mentioned Semrush, despite the company's blog content being cited hundreds of times by the same platform. That earlier disclosure illustrated the same mention-versus-citation gap that the 2026 Index now quantifies at scale across more than 1,200 brands.

Adobe completed its $1.9 billion acquisition of Semrush in November 2025, a deal that combined Adobe's customer experience tools with Semrush's search and generative engine optimization data. The AI Visibility Index 2026 represents one of the first major joint research outputs published under that combined ownership, with contributions in the report's foreword from Rachel Thornton, chief marketing officer of Adobe Enterprise, and Andrew Warden, vice president of marketing at Adobe and former chief marketing officer of Semrush.

The report's findings on citation volatility also track with patterns PPC Land has covered involving individual platforms. Research published in August 2025 found that ChatGPT referral traffic to publisher websites dropped 52% after OpenAI adjusted its citation weighting to favor Reddit and Wikipedia over branded and commercial sources, with Reddit citations rising 87% and Wikipedia citations rising 62% during the same window. The Semrush data released today shows Reddit and Wikipedia continuing to function as dominant citation sources across multiple platforms nearly a year later, with Reddit cited 77.4 million times in the Community cluster analysis and Wikipedia cited 71.7 million times in the Education & Training vertical alone.

Separate Semrush research published in June 2026 examined how 89,000 LinkedIn URLs are cited across ChatGPT Search, Google AI Mode and Perplexity, finding that LinkedIn ranked as the second most-cited domain in AI search responses at an 11% citation rate, trailing only Reddit. That earlier study, which focused specifically on business-to-business content strategy, complements the broader industry-wide patterns the AI Visibility Index 2026 now documents.

What the data does not resolve

The report itself acknowledges several limitations. Its Brand Leaderboard captures only the top 30 brands per platform, vertical and month, meaning brands ranked lower may not appear in strict cross-platform intersections such as the Universal 36. Cross-platform tile metrics rely on platform-native worldwide aggregates rather than country-specific breakdowns, while certain case study figures, including the headline Patagonia and Shopify numbers, are sourced from United States data specifically. The report also notes that brand entity resolution was not normalized across variants, meaning related domains and brand names belonging to the same parent company, such as Google and YouTube, or Chase and JPMorgan Chase, appear as separate entries in the dataset rather than consolidated figures.

The four AI platforms measured also update their underlying models and source-weighting periodically, a dynamic the report flags as a known limitation capable of shifting brand and source rankings between any two months in the study window, independent of any change in the underlying brand's own marketing activity.

Timeline

  • September 2025: Semrush's US AI search prompt database reaches 89 million deduplicated prompts, the starting point cited in the report's methodology.
  • January 2026 to April 2026: Semrush collects the four-month dataset of 126 million AI search prompts across ChatGPT, Google Gemini, Google AI Mode and Google AI Overviews that forms the basis of the report.
  • April 2026: Semrush's prompt database reaches 126 million unique prompts, the final figure cited throughout the report.
  • June 24, 2026: ZECOMMS Agency distributes an embargoed briefing on the AI Visibility Index 2026 to media ahead of the planned release.
  • June 25, 2026: The originally scheduled public release date for the AI Visibility Index 2026.
  • June 29, 2026: The report goes live publicly at ai-visibility-index.semrush.com, according to follow-up correspondence from ZECOMMS Agency.
  • June 30, 2026: Semrush, an Adobe company, publishes the AI Visibility Index 2026 in full.

Summary

Who: Semrush, the search and generative engine optimization platform acquired by Adobe in November 2025, conducted the research. Contributors to the report include Rachel Thornton, chief marketing officer of Adobe Enterprise, and Andrew Warden, vice president of marketing at Adobe and former chief marketing officer of Semrush, alongside six external marketing practitioners interviewed for the report's expert voices chapter.

What: Semrush published the AI Visibility Index 2026, an analysis of 126 million United States AI search prompts examining how more than 1,200 brands across 22 industry verticals are mentioned and cited by ChatGPT, Google Gemini, Google AI Mode and Google AI Overviews. The report found that only 36 brands, termed the Universal 36, achieved top-100 visibility on every platform in every month of the study window, while mention-source overlap, the alignment between brands an AI names and the sources it quotes, ranged from 64% on Google AI Overviews to as low as 30% on Gemini.

When: The underlying dataset covers January through April 2026. An embargoed briefing was distributed to media on June 24, 2026, ahead of a planned June 25 release; the report went publicly live on June 29, 2026, and Semrush published it in full today, June 30, 2026.

Where: The dataset is built from United States AI search prompts. The report is published by Semrush, headquartered in Boston, operating as an Adobe company following the November 2025 acquisition.

Why: The report matters to the marketing community because it quantifies, at a scale of more than 1,200 brands, a measurement gap that a separate Semrush survey found 45% of marketers cannot currently resolve on their own. The finding that mentions and citations move independently, sometimes by a margin of 34 percentage points depending on the platform, has direct implications for how marketing, content, public relations and customer experience teams divide responsibility for a brand's presence in AI-generated answers, a division of labor the report argues most organizations have not yet established.