Earned media and AI citations: 84% holds broadly, but brand sites lead on product questions
Earned media supplies 84% of AI citations in a 25-million-link study, but the share depends on the question. On factual pricing and product questions, 63.4% of citations in a 20,000-prompt test went to official websites. The circulating 82-95% earned-media figure is overstated as a general rule.
A figure circulating in PR-agency and vendor posts says 82 to 95% of AI citations come from earned media, not owned content. The evidence rates that claim Overstated. The lower end is real: a 25-million-link study found earned media behind 84% of AI citations Zebian · Muck Rack (Generative Pulse) · 2026. The share falls sharply once the question is about a brand's own product. On factual questions about pricing, features or company details, 63.4% of citations went to official websites Linehan · Ahrefs · 2026. The sections below show why the two numbers measure different things.
The 84% earned media share of AI citations is real, and the category is broad
The 84% figure comes from one vendor study with a large sample. Muck Rack analysed more than 25 million links cited by ChatGPT, Claude and Gemini across 17 industries Zebian · Muck Rack (Generative Pulse) · 2026. The earned share held between 82% and 89% across three editions since July 2025. Paid and advertorial content took 0.3%.
The study's "earned" is much wider than press coverage. Journalism accounts for 27 of the 84 points Zebian · Muck Rack (Generative Pulse) · 2026. The remaining 57 points are other non-owned sources, and the study does not publish an owned share for comparison. An academic analysis of more than 366,000 citations found news made up only 9% of all citations Yang · arXiv (cs.IR) · 2025.
The upper end of the circulating range has no traceable study behind it. No dataset in this library reports 95% earned media as a general figure. The 84% is a defensible headline for broad prompt sets, with one vendor's definition attached.
The owned share of AI citations depends on the question asked
Brand-owned pages win some question types and lose others. Omniscient Digital ran 240 branded prompts across five AI engines and classified 23,387 cited sources Dombrowski · Omniscient Digital · 2026. Owned content was 23% overall, earned media 48% and other brands' content 30%. The study is one agency's work with author-selected prompts.
The intent split is the useful part. Owned content led on functionality and integration questions, at 50% Dombrowski · Omniscient Digital · 2026. Earned media dominated customer review questions, at 82%. A brand's own site answers what the product does. Third parties answer whether it is any good.
A larger test points the same way. In 20,000 US prompts, 63.4% of citations on factual pricing, feature and company questions went to official websites, against 46.4% on comparison questions Linehan · Ahrefs · 2026. The test covers one free, logged-out ChatGPT model snapshot. It shows that the owned share rises when the answer lives on the brand's site.
How AI citations are counted decides the earned versus owned split
Business listings show how much the label matters. Yext sorted 6.8 million citations by how much control a brand has: websites supplied 44% and listings 42% Yext · 2025. Yext counts listings as controllable, so 86% of citations come from sources brands can manage. Yext sells listings management, and its 2025 sample drew on its own clients across four local-search industries.
The denominator matters as much as the label. Across 339 Indian brands, the median brand's own website supplied 0.13% of its category's AI citations Shaikh · Cited (getcited.in) · 2026. All brand websites pooled were still the largest source, at 54.8%, and reached 84% to 90% in B2B software categories. One brand's site is small; brand sites as a class are not.
Agency data across eight brands found own-domain share ranging from 0% to 74.7% Kidd · DeltaV Digital · 2026. Own domains that were retrieved had the highest citation rate of any domain type. The sample is small, but it locates the gap in retrieval rather than in trust.
Why AI cites Wikipedia, Reddit and YouTube: general queries, not buying questions
The domain-concentration findings come from broad, mostly informational prompt sets. Across 46 million Google AI Overview citations, YouTube took about 23.3%, Wikipedia 18.4% and Google.com 16.4% Surfer SEO · 2025. Wikipedia was ChatGPT's top source in a 680-million-citation panel, at 7.8% of its citations Profound (TryProfound) · 2025.
Commercial questions draw on a different web. Across about 16,400 commercial-intent answers on five engines, Reddit took 1.8% of citations and Wikipedia under 0.6% Khallad · SurfacedBy · 2026. Vendor, product and long-tail sites took 90.6%. The vendor does not say how many of those pages belonged to the brand in question.
Reddit's place in AI answers is also unstable. ChatGPT's Reddit citations fell from close to 60% of responses to around 10% in September 2025 Semrush · 2025. A second drop in August 2026 took Reddit from 3.83% to 0.52% of ChatGPT Search citations Foppen · Promptwatch B.V. · 2026. The tracker flagged that figure as provisional. Perplexity's Reddit citations rose 93% over the same weeks Shaikh · Pierview · 2026.
Third-party mentions still track AI visibility, and competitors fill the gap
The overstatement does not make earned media irrelevant. Across 75,000 brands, branded web mentions correlated with AI Overview visibility at 0.664, against 0.218 for backlinks Linehan · Ahrefs · 2025. That is correlation, and strong brands earn both mentions and visibility.
When a brand is absent from third-party sources, its competitors often speak for it. Around 40% of the outside sources cited in AI answers about a brand were that brand's direct competitors Khallad · SurfacedBy · 2026. The study comes from a tracking vendor with commercial interest and covers about 100,000 citations.
Review platforms show the same pattern with a large selection effect. Paid G2 profiles earned a median of 806 AI citations over 180 days, against 8 for free listings Indig · G2 · 2026. Within matched review-count buckets the gap shrinks to roughly 2x to 7x. G2 sells the paid profiles it studied.
FAQFrequently Asked Questions
Sources
Sources are tiered per our methodology & sources page.
News Source Citing Patterns in AI Search Systems
arXiv (cs.IR) · Kai-Cheng Yang · 2025
AI search systems concentrate news citations in a small set of outlets, and the cited mix leans politically liberal. Low-credibility sources are rarely cited. News makes up only 9% of all citations across more than 366,000 citations studied, so brands depending on press coverage for AI visibility face a narrow set of gatekeeper publishers, with limited influence from political leaning or quality on user satisfaction.
Methodology note
Academic analysis of the AI Search Arena platform, covering more than 24,000 conversations and 65,000 responses from search systems by OpenAI, Perplexity, and Google. The study extracted over 366,000 citations, isolated those referencing news, and correlated source-level attributes (political leaning, credibility ratings) with user preference data from head-to-head model comparisons.
An Analysis of AI Overview Brand Visibility Factors (75K Brands Studied)
Ahrefs · Louise Linehan, Xibeijia Guan · 2025
Across 75,000 brands, branded web mentions had the strongest Spearman correlation with AI Overview brand visibility (0.664), ahead of branded anchors (0.527), branded search volume (0.392), Domain Rating (0.326) and backlinks (0.218), so mentions correlate about 3x more strongly than backlinks. The top quartile by web mentions had a median of 169 AIO mentions against 14 for the next; 26% of brands had none.
Methodology note
Single-vendor Ahrefs study: 75,000 brands filtered to Domain Rating above 40 and a top keyword of at least 800 monthly searches, with AI Overview mentions measured through Ahrefs Brand Radar and Spearman rank correlation. Metrics are Ahrefs-defined and the authors stress correlation is not causation and all factors are moderate to weak. Direction corroborated by Seer Interactive and Kevin Indig.
How ChatGPT Does Query Fan-Out (15 Million Queries Studied) (Ahrefs)
Ahrefs · Louise Linehan · 2026
Across 66.6 million logged-out GPT-5.6 Luna responses and 15 million fan-out queries (August to September 2026), only 14.4% of responses contained fan-out queries and 6.8% of fan-outs used the site: operator. Among the roughly 160 most-searched sites, 75.9% of site: searches targeted brand sites. Reddit appeared in about 1 in 5 US answers in March but about 1 in 43 by late September.
Methodology note
Single-vendor Ahrefs study by Louise Linehan, reviewed by Ryan Law, published 12 October 2026 and fetched directly. Uses 66.6 million free logged-out responses, Brand Radar US citations from March to September 2026 (2.4 to 3.8 million answers a month), a 20,000-prompt official-source test and about one million site: queries. The article limits findings to one free-tier model snapshot.
Reddit Citations Fell Off a Cliff in ChatGPT, While Perplexity Doubled Down (Pierview)
Pierview · Naqui Shaikh · 2026
Across 44,488 tracked Reddit citations from 1 July to 19 August 2026, ChatGPT's daily Reddit citations fell from about 248 in early August to 16 over 14-19 August, a 94% collapse in absolute counts, not share. In the same window Perplexity's rose 93% (512 to 988 per day) and supplied 71% of all tracked Reddit citations, while Gemini stayed roughly flat.
Methodology note
Vendor study by Pierview, an AI-visibility platform, authored by Naqui Shaikh and last updated 24 August 2026, fetched directly on 28 August 2026. Continuous real-browser session tracking of 50,000-plus prompts across six AI surfaces; Reddit citations counted daily per engine. Prompt-set composition, category mix and confidence intervals are not disclosed.
Why Did ChatGPT Stop Citing Reddit? Inside the August 2026 Citation Collapse (Promptwatch)
Promptwatch B.V. · Klaas Foppen · 2026
Reddit's share of ChatGPT Search citations averaged 3.83% from 18 July to 7 August 2026, then fell to 0.52% across 14-17 August, an 86.4% relative drop. Over the same window Google AI Overviews fell only 11.3% (2.37% to 2.10%) and AI Mode 30.5% (2.22% to 1.54%), both gradually. Promptwatch calls the drop provisional and cannot rule out its own collection error.
Methodology note
First-party vendor telemetry from Promptwatch, an AI-visibility tracker, authored by co-founder Klaas Foppen and published 20 August 2026, fetched directly on 28 August 2026. Reddit's share is computed over ChatGPT Search responses containing at least one citation. No prompt-panel size, panel composition or confidence intervals are disclosed.
AI Is Citing Your Brand From Everywhere Except Your Own Website (Cited Index, June 2026)
Cited (getcited.in) · Salman Shaikh · 2026
Across 339 Indian brands on five AI platforms, the median brand's own website supplied 0.13% of its category's AI citations (mean 0.65%), yet all brand websites pooled were the largest source at 54.8%. Brand sites carried 84 to 90% of citations in B2B software categories such as HR and payroll and CRM, but 27 to 62% in consumer categories. Single-vendor index, India only.
Methodology note
Blog post by Salman Shaikh of Cited (getcited.in), an Indian AI-visibility vendor, published 10 August 2026, summarising its bi-monthly Cited Index across ChatGPT, Perplexity, Gemini, AI Overviews and AI Mode. The methodology page, read 2026-10-06, discloses 19 to 29 prompts per category run once per platform. Page fetched directly.
AI Search Citations Study: What 25,000+ Citations Reveal (DeltaV Digital 2026)
DeltaV Digital · Brandon Kidd · 2026
Citation mixes differ sharply by industry: listicles took 61% of AI citations in B2B technology services but 0% in healthcare, homepages 55% for a local-services brand, and official program pages 53% in higher education. Own-domain citation share ranged from 0% to 74.7%, yet retrieved own domains showed the highest citation rate of any domain type, suggesting the bottleneck is retrieval, not trust. Eight brands, directional.
Methodology note
First original-research release from agency DeltaV Digital, authored by Brandon Kidd (VP Operations), published 13 July 2026. 21,075 AI responses tracked daily via Peec AI from 14 April to 13 July 2026 across ChatGPT, Perplexity, Gemini, AI Overviews and AI Mode for eight anonymised client brands; top-30 cited URLs and domains per brand analysed (25,337 citations). Limitations self-disclosed. Verified by direct fetch.
When AI Describes Your Brand, It Cites Your Competitors (SurfacedBy)
SurfacedBy · Ali Khallad · 2026
SurfacedBy analyzed almost 100,000 citations to outside sources behind AI answers about a sample of brands, across ChatGPT, Claude, Gemini, Perplexity, and Google AI Mode over about three months in spring 2026. Around 40% of the outside sources cited about a brand were that brand's direct competitors; the typical brand sat above a third, and nearly nine in ten had at least a quarter. Brand-owned pages are a smaller stream.
Methodology note
First-party study by SurfacedBy, an AI-visibility tracking vendor with commercial interest, published 29 June 2026. Nearly 100,000 outside-source citations resolving to 10,000-plus domains, tagged competitor or third party using each brand's own competitor set, aggregated brand by brand. Authors stress a small, mixed brand sample; decimals provisional. Verified by direct fetch.
We Analyzed 127,198 AI Citations. The Five Engines Barely Read the Same Web. (SurfacedBy)
SurfacedBy · Ali Khallad · 2026
SurfacedBy analyzed 127,198 source citations from ChatGPT, Claude, Gemini, Perplexity, and Google AI Mode across roughly 16,400 commercial-intent answers between March and June 2026. Of 11,647 cited domains, 69.6% were cited by only one engine and just 2.7% by all five. Vendor, product, and long-tail pages drew 90.6% of citations; Reddit 1.8% and Wikipedia 0.6%. Gemini averaged 11.0 sources per answer, ChatGPT 3.7.
Methodology note
First-party experiment by SurfacedBy, an AI-visibility tracking vendor with commercial interest, published 27 June 2026 and updated 29 June. About 16,400 answers to real buyer and category questions across five engines; citations counted at the domain level. Authors disclose limits: commercial-query skew, citations are not clicks, engine behavior shifts. Verified by direct fetch.
Earned Media Still Drives 84% of AI Citations (Muck Rack, What Is AI Reading? May 2026)
Muck Rack (Generative Pulse) · Linda Zebian · 2026
Muck Rack's May 2026 "What Is AI Reading?" study analyzed more than 25 million links cited by ChatGPT, Claude, and Gemini across 17 industries and found earned media accounts for 84% of all AI citations, journalism alone 27%, and paid/advertorial just 0.3%. The earned share held at 82-89% across three editions since July 2025. Single-vendor study; treat as directional.
Methodology note
Single-vendor study by Muck Rack's Generative Pulse team, published 7 May 2026, authored by Linda Zebian. Analyzed 25M+ citation links across ChatGPT, Claude, and Gemini over 17 industries; third consecutive edition since July 2025. Page fetched and figures, per-platform citation rates, and top-cited domains confirmed directly.
Do Paid G2 Profiles Earn More AI Citations? (G2, Kevin Indig, 84,623 products)
G2 · Kevin Indig · 2026
Across 84,623 G2 products over 180 days, paid profiles had a median 806 AI citations against 8 for free listings, a 101x gap at the median and 15x at the mean. Paid profiles are 9% of products but capture 60% of citations. Within review-count buckets the paid advantage falls to roughly 2x to 7x; the study is correlational.
Methodology note
First-party study published by G2 on its sell hub on 27 April 2026, authored by Kevin Indig, using Profound citation tracking across ChatGPT, Perplexity, AI Overviews and AI Mode. Paid and free products are compared within review-count buckets. Page fetched directly. No uncertainty intervals or precise citation definition; G2 sells paid profiles.
How LLMs Source Brand Information: An Analysis of 23,000+ AI Citations (Omniscient Digital)
Omniscient Digital · Cate Dombrowski · 2026
Omniscient Digital analysed 23,387 sources from 240 branded prompts across ChatGPT, Perplexity, Gemini, AI Mode and AI Overviews: about 48% earned media, 30% commercial brand content from other brands, and 23% owned. Owned content led for functionality and integration queries (about 50%), while earned media dominated customer review queries (about 82%).
Methodology note
Agency research post by Cate Dombrowski, Omniscient Digital, published 8 January 2026. Prompts were split across four industries and six intents and run through Peec AI; the author classified citations with Python. Brand count and edge-case rules are not disclosed, and the agency sells GEO services. Read directly from the page.
Across more than 100 million citations from 230,000 prompts tracked weekly between July 14 and October 12, 2025, ChatGPT's reliance on Reddit and Wikipedia collapsed in mid-September. Reddit fell from close to 60% of ChatGPT responses in early August to around 10% by mid-September. Wikipedia dropped from roughly 55% to under 20%. AI Mode and Perplexity stayed stable; LinkedIn and Forbes citations grew across all three engines.
Methodology note
Semrush ran weekly snapshots of citations for 230,000 prompts over 13 weeks across ChatGPT search, Google AI Mode and Perplexity. Each week the team tracked the 25 most-cited domains and totaled changes around Google's mid-September removal of the num=100 search parameter. The post includes per-platform domain trend lines and lists the biggest gainers and losers per engine.
AI Citations, User Locations and Query Context: Where AI Models Get Their Brand Answers (Yext, 6.8M citations)
Yext · 2025
Yext analysed 6.8 million citations from about 1.6 million responses across Gemini, OpenAI and Perplexity, sorted by how much control a brand has. Websites supplied 44% and listings 42%, with reviews and social at 8% and uncontrollable sources at 6%. About 52.15% of Gemini citations came from first-party websites. Data is from 2025 and the sample is Yext clients and prospects.
Methodology note
Vendor research page by Yext, with the announcement dated 9 October 2025 on its blog; page date itself not shown. Citations were pulled through each model's API with Yext Scout, 1 July to 31 August 2025, across four industries and four branded or objective query quadrants. Page read directly on 30 October 2026; Yext sells listings management.
AI Platform Citation Patterns (680M citations across ChatGPT, AI Overviews, Perplexity)
Profound (TryProfound) · 2025
Across 680 million tracked citations between August 2024 and June 2025, the three big AI engines source very differently. Wikipedia is ChatGPT's top source at 7.8% of citations and 47.9% of its top-10 sources. Reddit leads on Perplexity (6.6% of all citations, 46.7% of top-10) and Google AI Overviews (2.2%). Around 80% of cited URLs sit on .com domains.
Methodology note
Profound analysed citations collected by its monitoring platform across ChatGPT, Google AI Overviews and Perplexity from August 2024 to June 2025. The post reports two cuts of the same data: share of total citations (per platform) and share of each platform's top 10 most-cited sources. Top-level domain distribution is broken out separately. Source lists are published in the article.
Surfer SEO AI Citation Report 2025 (36M AIOs, 46M citations)
Surfer SEO · 2025
Across 36 million Google AI Overviews and 46 million citations between March and August 2025, three domains dominate: YouTube at about 23.3%, Wikipedia at 18.4% and Google.com at 16.4%. Industry mix shifts the picture: NIH leads health at 39%, YouTube and Reddit together carry gaming with 93% and 78% appearance rates, and Shopify takes 17.7% of ecommerce citations.
Methodology note
Surfer's AI Tracker logged AI Overview responses and their citations from March to August 2025, covering 36M Overviews and 46M citations across 57,000-plus URLs. The team broke results into industry segments (finance, health, ecommerce, SEO, gaming, sports, travel) and reported the share of citations earned by the most frequent domains within each category.
About the author Max Ackermann
Max Ackermann is founder and Managing Director of info.link, the product data platform that makes brands visible in AI search and connects every physical product to the web through GS1 Digital Link. He writes about AI search and generative engine optimization (GEO), AI-powered commerce, and how brands can structure product data for ChatGPT, Gemini, Perplexity, and retailer AI assistants like Amazon Rufus. For the past two years he has built the pipelines that put structured product data into AI answers, and run the experiments that test what actually moves AI citations.
Max has 20+ years of experience building digital products and businesses. He previously led McKinsey's Corporate Venture and Design teams across Europe, and as Managing Director of a leading US digital agency he built platforms with Nike, Google, Meta, and Airbnb. He founded the UX Design program at Central Saint Martins College, University of the Arts London, and is a Fellow of the UK's Higher Education Academy. Based in Hamburg, he works closely with GS1 on Digital Link adoption; info.link is headquartered in Hamburg and Berlin and counts GS1 Germany among its investors.
Follow Max on LinkedIn.
